Daniel Fried

Loubna Ben Allal, Raymond Li, Denis Kocetkov et al. · cmu, huggingface

The BigCode project is an open-scientific collaboration working on the responsible development of large language models for code. This tech report describes the progress of the collaboration until December 2022, outlining the current state of the Personally Identifiable Information (PII) redaction pipeline, the experiments conducted to de-risk the model architecture, and the experiments investigating better preprocessing methods for the training data. We train 1.1B parameter models on the Java, JavaScript, and Python subsets of The Stack and evaluate them on the MultiPL-E text-to-code benchmark. We find that more aggressive filtering of near-duplicates can further boost performance and, surprisingly, that selecting files from repositories with 5+ GitHub stars deteriorates performance significantly. Our best model outperforms previous open-source multilingual code generation models (InCoder-6.7B and CodeGen-Multi-2.7B) in both left-to-right generation and infilling on the Java, JavaScript, and Python portions of MultiPL-E, despite being a substantially smaller model. All models are released under an OpenRAIL license at https://hf.co/bigcode.

Freda Shi, Daniel Fried, Marjan Ghazvininejad et al. · cmu, uw

Generative models of code, pretrained on large corpora of programs, have shown great success in translating natural language to code (Chen et al., 2021; Austin et al., 2021; Li et al., 2022, inter alia). While these models do not explicitly incorporate program semantics (i.e., execution results) during training, they are able to generate correct solutions for many problems. However, choosing a single correct program from a generated set for each problem remains challenging. In this work, we introduce execution result--based minimum Bayes risk decoding (MBR-EXEC) for program selection and show that it improves the few-shot performance of pretrained code models on natural-language-to-code tasks. We select output programs from a generated candidate set by marginalizing over program implementations that share the same semantics. Because exact equivalence is intractable, we execute each program on a small number of test inputs to approximate semantic equivalence. Across datasets, execution or simulated execution significantly outperforms the methods that do not involve program semantics. We find that MBR-EXEC consistently improves over all execution-unaware selection methods, suggesting it as an effective approach for natural language to code translation. We open-source our code at github.com/facebookresearch/mbr-exec and data at dl.fbaipublicfiles.com/mbr-exec/mbr-exec-release.zip

Grace Luo, Giscard Biamby, Trevor Darrell et al. · cmu

We demonstrate how language can improve geolocation: the task of predicting the location where an image was taken. Here we study explicit knowledge from human-written guidebooks that describe the salient and class-discriminative visual features humans use for geolocation. We propose the task of Geolocation via Guidebook Grounding that uses a dataset of StreetView images from a diverse set of locations and an associated textual guidebook for GeoGuessr, a popular interactive geolocation game. Our approach predicts a country for each image by attending over the clues automatically extracted from the guidebook. Supervising attention with country-level pseudo labels achieves the best performance. Our approach substantially outperforms a state-of-the-art image-only geolocation method, with an improvement of over 5% in Top-1 accuracy. Our dataset and code can be found at https://github.com/g-luo/geolocation_via_guidebook_grounding.

Daniel Fried, Armen Aghajanyan, Jessy Lin et al. · berkeley, cmu

Code is seldom written in a single left-to-right pass and is instead repeatedly edited and refined. We introduce InCoder, a unified generative model that can perform program synthesis (via left-to-right generation) as well as editing (via infilling). InCoder is trained to generate code files from a large corpus of permissively licensed code, where regions of code have been randomly masked and moved to the end of each file, allowing code infilling with bidirectional context. Our model is the first generative model that is able to directly perform zero-shot code infilling, which we evaluate on challenging tasks such as type inference, comment generation, and variable re-naming. We find that the ability to condition on bidirectional context substantially improves performance on these tasks, while still performing comparably on standard program synthesis benchmarks in comparison to left-to-right only models pretrained at similar scale. The InCoder models and code are publicly released. https://sites.google.com/view/incoder-code-models

Shuyan Zhou, Frank F. Xu, Hao Zhu et al. · cmu

With advances in generative AI, there is now potential for autonomous agents to manage daily tasks via natural language commands. However, current agents are primarily created and tested in simplified synthetic environments, leading to a disconnect with real-world scenarios. In this paper, we build an environment for language-guided agents that is highly realistic and reproducible. Specifically, we focus on agents that perform tasks on the web, and create an environment with fully functional websites from four common domains: e-commerce, social forum discussions, collaborative software development, and content management. Our environment is enriched with tools (e.g., a map) and external knowledge bases (e.g., user manuals) to encourage human-like task-solving. Building upon our environment, we release a set of benchmark tasks focusing on evaluating the functional correctness of task completions. The tasks in our benchmark are diverse, long-horizon, and designed to emulate tasks that humans routinely perform on the internet. We experiment with several baseline agents, integrating recent techniques such as reasoning before acting. The results demonstrate that solving complex tasks is challenging: our best GPT-4-based agent only achieves an end-to-end task success rate of 14.41%, significantly lower than the human performance of 78.24%. These results highlight the need for further development of robust agents, that current state-of-the-art large language models are far from perfect performance in these real-life tasks, and that WebArena can be used to measure such progress.

Xiang Lisa Li, Ari Holtzman, Daniel Fried et al. · cmu, microsoft-research

Given a language model (LM), maximum probability is a poor decoding objective for open-ended generation, because it produces short and repetitive text. On the other hand, sampling can often produce incoherent text that drifts from the original topics. We propose contrastive decoding (CD), a reliable decoding approach that optimizes a contrastive objective subject to a plausibility constraint. The contrastive objective returns the difference between the likelihood under a large LM (called the expert, e.g. OPT-13B) and a small LM (called the amateur, e.g. OPT-125M), and the constraint ensures that the outputs are plausible. CD is inspired by the fact that the failures of larger LMs (e.g., repetition, incoherence) are even more prevalent in smaller LMs, and that this difference signals which texts should be preferred. CD requires zero additional training, and produces higher quality text than decoding from the larger LM alone. It also works across model scales (OPT-13B and GPT2-1.5B) and significantly outperforms four strong decoding algorithms (e.g., nucleus, top-k) in automatic and human evaluations across wikipedia, news and story domains.

Yuhang Lai, Chengxi Li, Yiming Wang et al. · cmu, uw

We introduce DS-1000, a code generation benchmark with a thousand data science problems spanning seven Python libraries, such as NumPy and Pandas. Compared to prior works, DS-1000 incorporates three core features. First, our problems reflect diverse, realistic, and practical use cases since we collected them from StackOverflow. Second, our automatic evaluation is highly specific (reliable) -- across all Codex-002-predicted solutions that our evaluation accept, only 1.8% of them are incorrect; we achieve this with multi-criteria metrics, checking both functional correctness by running test cases and surface-form constraints by restricting API usages or keywords. Finally, we proactively defend against memorization by slightly modifying our problems to be different from the original StackOverflow source; consequently, models cannot answer them correctly by memorizing the solutions from pre-training. The current best public system (Codex-002) achieves 43.3% accuracy, leaving ample room for improvement. We release our benchmark at https://ds1000-code-gen.github.io.

Maarten Sap, Ronan LeBras, Daniel Fried et al. · allen-ai, cmu

Social intelligence and Theory of Mind (ToM), i.e., the ability to reason about the different mental states, intents, and reactions of all people involved, allow humans to effectively navigate and understand everyday social interactions. As NLP systems are used in increasingly complex social situations, their ability to grasp social dynamics becomes crucial. In this work, we examine the open question of social intelligence and Theory of Mind in modern NLP systems from an empirical and theory-based perspective. We show that one of today's largest language models (GPT-3; Brown et al., 2020) lacks this kind of social intelligence out-of-the box, using two tasks: SocialIQa (Sap et al., 2019), which measures models' ability to understand intents and reactions of participants of social interactions, and ToMi (Le et al., 2019), which measures whether models can infer mental states and realities of participants of situations. Our results show that models struggle substantially at these Theory of Mind tasks, with well-below-human accuracies of 55% and 60% on SocialIQa and ToMi, respectively. To conclude, we draw on theories from pragmatics to contextualize this shortcoming of large language models, by examining the limitations stemming from their data, neural architecture, and training paradigms. Challenging the prevalent narrative that only scale is needed, we posit that person-centric NLP approaches might be more effective towards neural Theory of Mind. In our updated version, we also analyze newer instruction tuned and RLFH models for neural ToM. We find that even ChatGPT and GPT-4 do not display emergent Theory of Mind; strikingly even GPT-4 performs only 60% accuracy on the ToMi questions related to mental states and realities.

Xuhui Zhou, Hao Zhu, Leena Mathur et al. · allen-ai, cmu

Humans are social beings; we pursue social goals in our daily interactions, which is a crucial aspect of social intelligence. Yet, AI systems' abilities in this realm remain elusive. We present SOTOPIA, an open-ended environment to simulate complex social interactions between artificial agents and evaluate their social intelligence. In our environment, agents role-play and interact under a wide variety of scenarios; they coordinate, collaborate, exchange, and compete with each other to achieve complex social goals. We simulate the role-play interaction between LLM-based agents and humans within this task space and evaluate their performance with a holistic evaluation framework called SOTOPIA-Eval. With SOTOPIA, we find significant differences between these models in terms of their social intelligence, and we identify a subset of SOTOPIA scenarios, SOTOPIA-hard, that is generally challenging for all models. We find that on this subset, GPT-4 achieves a significantly lower goal completion rate than humans and struggles to exhibit social commonsense reasoning and strategic communication skills. These findings demonstrate SOTOPIA's promise as a general platform for research on evaluating and improving social intelligence in artificial agents.

Jing Yu Koh, Ruslan Salakhutdinov, Daniel Fried · cmu

We propose an efficient method to ground pretrained text-only language models to the visual domain, enabling them to process arbitrarily interleaved image-and-text data, and generate text interleaved with retrieved images. Our method leverages the abilities of language models learnt from large scale text-only pretraining, such as in-context learning and free-form text generation. We keep the language model frozen, and finetune input and output linear layers to enable cross-modality interactions. This allows our model to process arbitrarily interleaved image-and-text inputs, and generate free-form text interleaved with retrieved images. We achieve strong zero-shot performance on grounded tasks such as contextual image retrieval and multimodal dialogue, and showcase compelling interactive abilities. Our approach works with any off-the-shelf language model and paves the way towards an effective, general solution for leveraging pretrained language models in visually grounded settings.

Zhiruo Wang, Shuyan Zhou, Daniel Fried et al. · cmu

To extend the scope of coding queries to more realistic settings, we propose ODEX, the first Open-Domain EXecution-based natural language (NL) to Python code generation dataset. ODEX has 945 NL-Code pairs spanning 79 diverse libraries, along with 1,707 human-written test cases for execution. Our NL-Code pairs are harvested from StackOverflow forums to encourage natural and practical coding queries. Moreover, ODEX supports four natural languages as intents, in English, Spanish, Japanese, and Russian. ODEX unveils intriguing behavioral differences among top-performing code language models (LM). While CODEX achieves better overall results, CODEGEN improves effectively via scaling -- CODEGEN 6.1B performs comparably with CODEX 12B. Both models show substantial gaps between open and closed domains, but CODEGEN gaps tend to decrease with model size while CODEX gaps increase. We release ODEX to facilitate research into open-domain problems for the code generation community.

Jessy Lin, Daniel Fried, Dan Klein et al. · cmu

In classic instruction following, language like "I'd like the JetBlue flight" maps to actions (e.g., selecting that flight). However, language also conveys information about a user's underlying reward function (e.g., a general preference for JetBlue), which can allow a model to carry out desirable actions in new contexts. We present a model that infers rewards from language pragmatically: reasoning about how speakers choose utterances not only to elicit desired actions, but also to reveal information about their preferences. On a new interactive flight-booking task with natural language, our model more accurately infers rewards and predicts optimal actions in unseen environments, in comparison to past work that first maps language to actions (instruction following) and then maps actions to rewards (inverse reinforcement learning).

Tianyi Zhang, Tao Yu, Tatsunori B. Hashimoto et al. · cmu

Sampling diverse programs from a code language model and reranking with model likelihood is a popular method for code generation but it is prone to preferring degenerate solutions. Inspired by collaborative programming, we propose Coder-Reviewer reranking. We augment Coder language models from past work, which generate programs given language instructions, with Reviewer models, which evaluate the likelihood of the instruction given the generated programs. We perform an extensive study across six datasets with eight models from three model families. Experimental results show that Coder-Reviewer reranking leads to consistent and significant improvement (up to 17% absolute accuracy gain) over reranking with the Coder model only. When combined with executability filtering, Coder-Reviewer reranking can often outperform the minimum Bayes risk method. Coder-Reviewer reranking is easy to implement by prompting, can generalize to different programming languages, and works well with off-the-shelf hyperparameters.

Yiqing Xie, Atharva Naik, Daniel Fried et al. · cmu

One major challenge of translating code between programming languages is that parallel training data is often limited. To overcome this challenge, we present two data augmentation techniques, one that builds comparable corpora (i.e., code pairs with similar functionality), and another that augments existing parallel data with multiple reference translations. Specifically, we build and analyze multiple types of comparable corpora, including programs generated from natural language documentation using a code generation model. Furthermore, to reduce overfitting to a single reference translation, we automatically generate additional translation references for available parallel data and filter the translations by unit tests, which increases variation in target translations. Experiments show that our data augmentation techniques significantly improve CodeT5 for translation between Java, Python, and C++ by an average of 7.5% Computational Accuracy (CA@1), which verifies the correctness of translations by execution. The code is available at https://github.com/Veronicium/CMTrans.

Atharva Naik, Marcus Alenius, Daniel Fried et al. · cmu

The task of automated code review has recently gained a lot of attention from the machine learning community. However, current review comment evaluation metrics rely on comparisons with a human-written reference for a given code change (also called a diff). Furthermore, code review is a one-to-many problem, like generation and summarization, with many "valid reviews" for a diff. Thus, we develop CRScore - a reference-free metric to measure dimensions of review quality like conciseness, comprehensiveness, and relevance. We design CRScore to evaluate reviews in a way that is grounded in claims and potential issues detected in the code by LLMs and static analyzers. We demonstrate that CRScore can produce valid, fine-grained scores of review quality that have the greatest alignment with human judgment among open source metrics (0.54 Spearman correlation) and are more sensitive than reference-based metrics. We also release a corpus of 2.9k human-annotated review quality scores for machine-generated and GitHub review comments to support the development of automated metrics.

Jing Yu Koh, Stephen McAleer, Daniel Fried et al. · cmu

Autonomous agents powered by language models (LMs) have demonstrated promise in their ability to perform decision-making tasks such as web automation. However, a key limitation remains: LMs, primarily optimized for natural language understanding and generation, struggle with multi-step reasoning, planning, and using environmental feedback when attempting to solve realistic computer tasks. Towards addressing this, we propose an inference-time search algorithm for LM agents to explicitly perform exploration and multi-step planning in interactive web environments. Our approach is a form of best-first tree search that operates within the actual environment space, and is complementary with most existing state-of-the-art agents. It is the first tree search algorithm for LM agents that shows effectiveness on realistic web tasks. On the challenging VisualWebArena benchmark, applying our search algorithm on top of a GPT-4o agent yields a 39.7% relative increase in success rate compared to the same baseline without search, setting a state-of-the-art success rate of 26.4%. On WebArena, search also yields a 28.0% relative improvement over a baseline agent, setting a competitive success rate of 19.2%. Our experiments highlight the effectiveness of search for web agents, and we demonstrate that performance scales with increased test-time compute. We conduct a thorough analysis of our results to highlight improvements from search, limitations, and promising directions for future work. Our code and models are publicly released at https://jykoh.com/search-agents.

Weiyan Shi, Emily Dinan, Adi Renduchintala et al. · cmu

Existing approaches built separate classifiers to detect nonsense in dialogues. In this paper, we show that without external classifiers, dialogue models can detect errors in their own messages introspectively, by calculating the likelihood of replies that are indicative of poor messages. For example, if an agent believes its partner is likely to respond "I don't understand" to a candidate message, that message may not make sense, so an alternative message should be chosen. We evaluate our approach on a dataset from the game Diplomacy, which contains long dialogues richly grounded in the game state, on which existing models make many errors. We first show that hand-crafted replies can be effective for the task of detecting nonsense in applications as complex as Diplomacy. We then design AutoReply, an algorithm to search for such discriminative replies automatically, given a small number of annotated dialogue examples. We find that AutoReply-generated replies outperform handcrafted replies and perform on par with carefully fine-tuned large supervised models. Results also show that one single reply without much computation overheads can also detect dialogue nonsense reasonably well.

Daniel Fried, Nicholas Tomlin, Jennifer Hu et al. · cmu

People rely heavily on context to enrich meaning beyond what is literally said, enabling concise but effective communication. To interact successfully and naturally with people, user-facing artificial intelligence systems will require similar skills in pragmatics: relying on various types of context -- from shared linguistic goals and conventions, to the visual and embodied world -- to use language effectively. We survey existing grounded settings and pragmatic modeling approaches and analyze how the task goals, environmental contexts, and communicative affordances in each work enrich linguistic meaning. We present recommendations for future grounded task design to naturally elicit pragmatic phenomena, and suggest directions that focus on a broader range of communicative contexts and affordances.

Jiefu Ou, Benno Krojer, Daniel Fried · cmu, mila

We propose a simple yet effective and robust method for contrastive captioning: generating discriminative captions that distinguish target images from very similar alternative distractor images. Our approach is built on a pragmatic inference procedure that formulates captioning as a reference game between a speaker, which produces possible captions describing the target, and a listener, which selects the target given the caption. Unlike previous methods that derive both speaker and listener distributions from a single captioning model, we leverage an off-the-shelf CLIP model to parameterize the listener. Compared with captioner-only pragmatic models, our method benefits from rich vision language alignment representations from CLIP when reasoning over distractors. Like previous methods for discriminative captioning, our method uses a hyperparameter to control the tradeoff between the informativity (how likely captions are to allow a human listener to discriminate the target image) and the fluency of the captions. However, we find that our method is substantially more robust to the value of this hyperparameter than past methods, which allows us to automatically optimize the captions for informativity - outperforming past methods for discriminative captioning by 11% to 15% accuracy in human evaluations

Jing Yu Koh, Ruslan Salakhutdinov, Daniel Fried

Computer use agents (CUAs) today are primarily deployed as single serial agents. This setup is suboptimal for complex long-horizon tasks that benefit from task decomposition, parallel execution, and consistent re-planning based on new information. In this paper, we argue that we should instead move towards evaluating and building multi-agent computer use (MACU) systems. These systems, which emphasize planning and parallel execution, alleviate many of the shortcomings of single-agent CUAs. We propose a general multi-agent setup in which a manager model decomposes computer use tasks as a directed acyclic graph (DAG), encoding relevant dependencies and goals for subagents. At each iteration, the manager dispatches parallel CUA subagents to carry out nodes on the ready frontier of the DAG, and continuously revises the DAG (adding, canceling, or rewriting nodes) as new findings arrive from subagents. This design treats the partially observable environment of computer use as a first class challenge: information that downstream agents may not be able to re-observe are retained and passed forward through the manager and DAG structure. We demonstrate that MACU consistently improves over strong single-agent baselines by $3.4-25.5\%$ on desktop (OSWorld) and web navigation (Online-Mind2Web, WebTailBench, Odysseys) benchmarks, exhibits more favorable test-time scaling, and solves complex long-horizon tasks where single-agent CUAs get stuck. On Odysseys, a long-horizon web navigation benchmark, MACU improves average task completion wall-clock time by ${\sim} 1.5 \times$, demonstrating its efficacy in speeding up traditionally slow CUA pipelines. Our findings highlight that multi-agent coordination is a promising axis for scaling computer use agents to work productively for longer and more effectively. We release all code and interactive visualizations at https://jykoh.com/multi-agent-computer-use.

Zora Zhiruo Wang, Sanidhya Vijayvargiya, Aspen Chen et al. · cmu

AI agents are increasingly developed and evaluated on benchmarks relevant to human work, yet it remains unclear how representative these benchmarking efforts are of the labor market as a whole. In this work, we systematically study the relationship between agent development efforts and the distribution of real-world human work by mapping benchmark instances to work domains and skills. We first analyze 43 benchmarks and 72,342 tasks, measuring their alignment with human employment and capital allocation across all 1,016 real-world occupations in the U.S. labor market. We reveal substantial mismatches between agent development that tends to be programming-centric, and the categories in which human labor and economic value are concentrated. Within work areas that agents currently target, we further characterize current agent utility by measuring their autonomy levels, providing practical guidance for agent interaction strategies across work scenarios. Building on these findings, we propose three measurable principles for designing benchmarks that better capture socially important and technically challenging forms of work: coverage, realism, and granular evaluation.

Joongwon Kim, Wannan Yang, Kelvin Niu et al.

Test-time scaling has become a powerful way to improve large language models. However, existing methods are best suited to short, bounded outputs that can be directly compared, ranked or refined. Long-horizon coding agents violate this premise: each attempt produces an extended trajectory of actions, observations, errors, and partial progress taken by the agent. In this setting, the main challenge is no longer generating more attempts, but representing prior experience in a form that can be effectively selected from and reused. We propose a test-time scaling framework for agentic coding based on compact representations of rollout trajectories. Our framework converts each rollout into a structured summary that preserves its salient hypotheses, progress, and failure modes while discarding low-signal trace details. This representation enables two complementary forms of inference-time scaling. For parallel scaling, we introduce Recursive Tournament Voting (RTV), which recursively narrows a population of rollout summaries through small-group comparisons. For sequential scaling, we adapt Parallel-Distill-Refine (PDR) to the agentic setting by conditioning new rollouts on summaries distilled from prior attempts. Our method consistently improves the performance of frontier coding agents across SWE-Bench Verified and Terminal-Bench v2.0. For example, by using our method Claude-4.5-Opus improves from 70.9% to 77.6% on SWE-Bench Verified (mini-SWE-agent) and 46.9% to 59.1% on Terminal-Bench v2.0 (Terminus 1). Our results suggest that test-time scaling for long-horizon agents is fundamentally a problem of representation, selection, and reuse.

Alex Wilf, Alex Tianyi Xu, Paul Pu Liang et al.

In the era of large scale pretrained models, Knowledge Distillation (KD) serves an important role in transferring the wisdom of computationally heavy teacher models to lightweight, efficient student models while preserving performance. Traditional KD paradigms, however, assume readily available access to teacher models for frequent inference -- a notion increasingly at odds with the realities of costly, often proprietary, large scale models. Addressing this gap, our paper considers how to minimize the dependency on teacher model inferences in KD in a setting we term Few Teacher Inference Knowledge Distillation (FTI KD). We observe that prevalent KD techniques and state of the art data augmentation strategies fall short in this constrained setting. Drawing inspiration from educational principles that emphasize learning through comparison, we propose Comparative Knowledge Distillation (CKD), which encourages student models to understand the nuanced differences in a teacher model's interpretations of samples. Critically, CKD provides additional learning signals to the student without making additional teacher calls. We also extend the principle of CKD to groups of samples, enabling even more efficient learning from limited teacher calls. Empirical evaluation across varied experimental settings indicates that CKD consistently outperforms state of the art data augmentation and KD techniques.

Sedrick Keh, Justin T. Chiu, Daniel Fried · cmu

When a model is trying to gather information in an interactive setting, it benefits from asking informative questions. However, in the case of a grounded multi-turn image identification task, previous studies have been constrained to polar yes/no questions, limiting how much information the model can gain in a single turn. We present an approach that formulates more informative, open-ended questions. In doing so, we discover that off-the-shelf visual question answering (VQA) models often make presupposition errors, which standard information gain question selection methods fail to account for. To address this issue, we propose a method that can incorporate presupposition handling into both question selection and belief updates. Specifically, we use a two-stage process, where the model first filters out images which are irrelevant to a given question, then updates its beliefs about which image the user intends. Through self-play and human evaluations, we show that our method is successful in asking informative open-ended questions, increasing accuracy over the past state-of-the-art by 14%, while resulting in 48% more efficient games in human evaluations.

Zora Zhiruo Wang, Jiayuan Mao, Daniel Fried et al.

Despite the potential of language model-based agents to solve real-world tasks such as web navigation, current methods still struggle with long-horizon tasks with complex action trajectories. In contrast, humans can flexibly solve complex tasks by learning reusable task workflows from past experiences and using them to guide future actions. To build agents that can similarly benefit from this process, we introduce Agent Workflow Memory (AWM), a method for inducing commonly reused routines, i.e., workflows, and selectively providing workflows to the agent to guide subsequent generations. AWM flexibly applies to both offline and online scenarios, where agents induce workflows from training examples beforehand or from test queries on the fly. We experiment on two major web navigation benchmarks -- Mind2Web and WebArena -- that collectively cover 1000+ tasks from 200+ domains across travel, shopping, and social media, among others. AWM substantially improves the baseline results by 24.6% and 51.1% relative success rate on Mind2Web and WebArena while reducing the number of steps taken to solve WebArena tasks successfully. Furthermore, online AWM robustly generalizes in cross-task, website, and domain evaluations, surpassing baselines from 8.9 to 14.0 absolute points as train-test task distribution gaps widen.

Yiqing Xie, Emmy Liu, Gaokai Zhang et al.

When assessing the quality of coding agents, predominant benchmarks focus on solving single issues on GitHub, such as SWE-Bench. In contrast, in real use, these agents solve more various and complex tasks that involve other skills such as exploring codebases, testing software, and designing architecture. In this paper, we first characterize some transferable skills that are shared across diverse tasks by decomposing trajectories into fine-grained components, and derive a set of principles for designing auxiliary training tasks to teach language models these skills. Guided by these principles, we propose a training environment, Hybrid-Gym, consisting of a set of scalable synthetic tasks, such as function localization and dependency search. Experiments show that agents trained on our synthetic tasks effectively generalize to diverse real-world tasks that are not present in training, improving a base model by 25.4% absolute gain on SWE-Bench Verified, 7.9% on SWT-Bench Verified, and 5.1% on Commit-0 Lite. Hybrid-Gym also complements datasets built for the downstream tasks (e.g., improving SWE-Play by 4.9% on SWT-Bench Verified). Code available at: https://github.com/yiqingxyq/Hybrid-Gym.

Andre He, Sean Welleck, Daniel Fried

Discrete diffusion models have recently become competitive with autoregressive models for language modeling, even outperforming them on reasoning tasks requiring planning and global coherence, but they require more computation at inference time. We trace this trade-off to a key mechanism: diffusion models are trained to jointly predict a distribution over all unknown tokens, including those that will not actually be decoded in the current step. Ablating this joint prediction yields faster inference but degrades performance, revealing that accurate prediction at the decoded position relies on joint reasoning about the distribution of undecoded tokens. We interpret these as latent tokens and introduce a method for modulating their number, demonstrating empirically that this enables a smooth tradeoff between inference speed and sample quality. Furthermore, we demonstrate that latent tokens can be introduced into autoregressive models through an auxiliary multi-token prediction objective, yielding substantial improvements on the same reasoning tasks where they have traditionally struggled. Our results suggest that latent tokens, while arising naturally in diffusion, represent a general mechanism for improving performance on tasks requiring global coherence or lookahead.

Yihan Cao, Shuyi Chen, Ryan Liu et al.

A persistent challenge to table question answering (TableQA) by generating executable programs has been adapting to varied table structures, typically requiring domain-specific logical forms. In response, this paper introduces a unified TableQA framework that: (1) provides a unified representation for structured tables as multi-index Pandas data frames, (2) uses Python as a powerful querying language, and (3) uses few-shot prompting to translate NL questions into Python programs, which are executable on Pandas data frames. Furthermore, to answer complex relational questions with extended program functionality and external knowledge, our framework allows customized APIs that Python programs can call. We experiment with four TableQA datasets that involve tables of different structures -- relational, multi-table, and hierarchical matrix shapes -- and achieve prominent improvements over past state-of-the-art systems. In ablation studies, we (1) show benefits from our multi-index representation and APIs over baselines that use only an LLM, and (2) demonstrate that our approach is modular and can incorporate additional APIs.

Lawrence Keunho Jang, Jing Yu Koh, Daniel Fried et al.

Existing web agent benchmarks have largely converged on short, single-site tasks that frontier models are approaching saturation on. However, real world web use consists of long-horizon, multi-site workflows. Common web navigation tasks, such as comparing products across different domains, planning trips across multiple services, or summarizing information from multiple search queries, require sustained context and cross-site reasoning over potentially hours of browsing. To capture and evaluate such behaviors, we introduce Odysseys: a benchmark of 200 long-horizon web tasks derived from real world browsing sessions evaluated on the live Internet. We find that binary pass/fail evaluation is inadequate for long-horizon settings and introduce a rubric-based evaluation, annotating each Odysseys task with an average of 6.1 graded rubrics. We demonstrate that this yields higher agreement with humans and provides a more fine-grained signal than commonly used trajectory-level LLM-as-a-judge evaluation metrics. We tested several leading frontier models and find that the strongest models achieve a success rate of 44.5%, which leaves substantial room for future improvements. Beyond task success, we argue that efficiency is a first-class concern for long-horizon agents. We introduce a Trajectory Efficiency metric (rubric score per step) and find that even frontier agents achieve only 1.15%, marking an evident need for agents that can succeed efficiently and not simply eventually. Odysseys isolates the critical evaluation of long-horizon proficiency in open-web environments, providing a realistic benchmark to measure progress towards computer-use agents that can potentially productively operate for hours. We release our tasks, evaluation scripts, and other results at https://odysseys-website.pages.dev

Yuxiang Wei, Olivier Duchenne, Jade Copet et al.

The recent DeepSeek-R1 release has demonstrated the immense potential of reinforcement learning (RL) in enhancing the general reasoning capabilities of large language models (LLMs). While DeepSeek-R1 and other follow-up work primarily focus on applying RL to competitive coding and math problems, this paper introduces SWE-RL, the first approach to scale RL-based LLM reasoning for real-world software engineering. Leveraging a lightweight rule-based reward (e.g., the similarity score between ground-truth and LLM-generated solutions), SWE-RL enables LLMs to autonomously recover a developer's reasoning processes and solutions by learning from extensive open-source software evolution data -- the record of a software's entire lifecycle, including its code snapshots, code changes, and events such as issues and pull requests. Trained on top of Llama 3, our resulting reasoning model, Llama3-SWE-RL-70B, achieves a 41.0% solve rate on SWE-bench Verified -- a human-verified collection of real-world GitHub issues. To our knowledge, this is the best performance reported for medium-sized (<100B) LLMs to date, even comparable to leading proprietary LLMs like GPT-4o. Surprisingly, despite performing RL solely on software evolution data, Llama3-SWE-RL has even emerged with generalized reasoning skills. For example, it shows improved results on five out-of-domain tasks, namely, function coding, library use, code reasoning, mathematics, and general language understanding, whereas a supervised-finetuning baseline even leads to performance degradation on average. Overall, SWE-RL opens up a new direction to improve the reasoning capabilities of LLMs through reinforcement learning on massive software engineering data.

Saujas Vaduguru, Daniel Fried, Yewen Pu

Programming-by-example is the task of synthesizing a program that is consistent with a set of user-provided input-output examples. As examples are often an under-specification of one's intent, a good synthesizer must choose the intended program from the many that are consistent with the given set of examples. Prior work frames program synthesis as a cooperative game between a listener (that synthesizes programs) and a speaker (a user choosing examples), and shows that models of computational pragmatic inference are effective in choosing the user intended programs. However, these models require counterfactual reasoning over a large set of programs and examples, which is infeasible in realistic program spaces. In this paper, we propose PraX, a novel way to amortize this search with neural networks. We sample pairs of programs and examples via self-play between listener and speaker models, and use pragmatic inference to choose informative training examples from this sample. We then use the informative dataset to train models to improve the synthesizer's ability to disambiguate user-provided examples without human supervision. We validate PraX on the challenging task of synthesizing regular expressions from example strings, and find that our method (1) outperforms models trained without choosing pragmatic examples by 23% (a 51% relative increase) (2) matches the performance of supervised learning on a dataset of pragmatic examples provided by humans, despite using no human data in training.

Siddhant Waghjale, Vishruth Veerendranath, Zora Zhiruo Wang et al.

Although large language models (LLMs) have been largely successful in generating functionally correct programs, conditioning models to produce efficient solutions while ensuring correctness remains a challenge. Further, unreliability in benchmarking code efficiency is a hurdle across varying hardware specifications for popular interpreted languages such as Python. In this paper, we present ECCO, a reproducible benchmark for evaluating program efficiency via two paradigms: natural language (NL) based code generation and history-based code editing. On ECCO, we adapt and thoroughly investigate the three most promising existing LLM-based approaches: in-context learning, iterative refinement with execution or NL feedback, and fine-tuning conditioned on execution and editing history. While most methods degrade functional correctness and moderately increase program efficiency, we find that adding execution information often helps maintain functional correctness, and NL feedback enhances more on efficiency. We release our benchmark to support future work on LLM-based generation of efficient code.

Justin T. Chiu, Wenting Zhao, Derek Chen et al.

Large language models (LLMs) excel at processing and generating both text and code. However, LLMs have had limited applicability in grounded task-oriented dialogue as they are difficult to steer toward task objectives and fail to handle novel grounding. We present a modular and interpretable grounded dialogue system that addresses these shortcomings by composing LLMs with a symbolic planner and grounded code execution. Our system consists of a reader and planner: the reader leverages an LLM to convert partner utterances into executable code, calling functions that perform grounding. The translated code's output is stored to track dialogue state, while a symbolic planner determines the next appropriate response. We evaluate our system's performance on the demanding OneCommon dialogue task, involving collaborative reference resolution on abstract images of scattered dots. Our system substantially outperforms the previous state-of-the-art, including improving task success in human evaluations from 56% to 69% in the most challenging setting.

Yuxiang Wei, Zhiqing Sun, Emily McMilin et al.

While current software agents powered by large language models (LLMs) and agentic reinforcement learning (RL) can boost programmer productivity, their training data (e.g., GitHub issues and pull requests) and environments (e.g., pass-to-pass and fail-to-pass tests) heavily depend on human knowledge or curation, posing a fundamental barrier to superintelligence. In this paper, we present Self-play SWE-RL (SSR), a first step toward training paradigms for superintelligent software agents. Our approach takes minimal data assumptions, only requiring access to sandboxed repositories with source code and installed dependencies, with no need for human-labeled issues or tests. Grounded in these real-world codebases, a single LLM agent is trained via reinforcement learning in a self-play setting to iteratively inject and repair software bugs of increasing complexity, with each bug formally specified by a test patch rather than a natural language issue description. On the SWE-bench Verified and SWE-Bench Pro benchmarks, SSR achieves notable self-improvement (+10.4 and +7.8 points, respectively) and consistently outperforms the human-data baseline over the entire training trajectory, despite being evaluated on natural language issues absent from self-play. Our results, albeit early, suggest a path where agents autonomously gather extensive learning experiences from real-world software repositories, ultimately enabling superintelligent systems that exceed human capabilities in understanding how systems are constructed, solving novel challenges, and autonomously creating new software from scratch.

Andre He, Daniel Fried, Sean Welleck

Reinforcement learning is emerging as a primary driver for improving language model reasoning capabilities. A fundamental question is whether current reinforcement learning algorithms -- such as Group Relative Policy Optimization (GRPO), the de facto standard algorithm used to improve language model reasoning -- merely sharpen the base model's distribution around problems it can already solve. We investigate this question in the context of formal theorem proving, which has access to a perfect verifier. We identify a degenerate rank bias in GRPO in which highly probable trajectories are reinforced and rare ones are neglected. This results in distribution sharpening: the model can solve some problems with fewer samples, but underperforms simply sampling more solutions from the original model. To overcome GRPO's rank bias we introduce unlikeliness reward, a simple method for explicitly up-weighting rare but correct solutions. We show that unlikeliness reward mitigates rank bias and improves pass@$N$ across a large range of $N$ in both synthetic and real theorem proving settings. We also uncover an unexpected link between rank bias and a seemingly mundane hyperparameter -- the number of updates per batch -- that leads to a second, complementary mitigation. We combine our insights into a revised GRPO training recipe for formal theorem proving, yielding an open pipeline that achieves competitive performance to DeepSeek-Prover-V1.5-RL on the miniF2F-test benchmark. We release our implementation at https://github.com/AndreHe02/rewarding-unlikely-release

Yiqing Xie, Alex Xie, Divyanshu Sheth et al.

To adequately test modern code generation systems, evaluation benchmarks must execute and test the code generated by the system. However, these execution and testing requirements have largely limited benchmarks to settings where code is easily executable or has human-written tests. To facilitate evaluation of code generation systems across diverse scenarios, we present CodeBenchGen, a framework to create scalable execution-based benchmarks from naturally occurring code sources. Specifically, we leverage a large language model (LLM) to sandbox arbitrary pieces of code into evaluation examples, including test cases for execution-based evaluation. We illustrate the usefulness of our framework by creating a dataset, Exec-CSN, which includes 1,931 examples involving 293 libraries converted from code in 367 GitHub repositories taken from the Code- SearchNet dataset. To demonstrate the solvability of examples in Exec-CSN, we present a human study demonstrating that 81.3% of the examples can be solved by humans and 61% are rated as "requires effort to solve". We conduct code generation experiments on open-source and proprietary models and analyze the performance of both humans and models. We provide code and data at: https://github.com/yiqingxyq/CodeBenchGen.

Gyeongwon James Kim, Alex Wilf, Louis-Philippe Morency et al.

Recent progress in autonomous code generation has fueled excitement around AI agents capable of accelerating scientific discovery by running experiments. However, there is currently no benchmark that evaluates whether such agents can implement scientific ideas when given varied amounts of code as a starting point, interpolating between reproduction (running code) and from-scratch replication (fully re-implementing and running code). We introduce AutoExperiment, a benchmark that evaluates AI agents' ability to implement and run machine learning experiments based on natural language descriptions in research papers. In each task, agents are given a research paper, a codebase with key functions masked out, and a command to run the experiment. The goal is to generate the missing code, execute the experiment in a sandboxed environment, and reproduce the results. AutoExperiment scales in difficulty by varying the number of missing functions $n$, ranging from partial reproduction to full replication. We evaluate state-of-the-art agents and find that performance degrades rapidly as $n$ increases. Agents that can dynamically interact with the environment (e.g. to debug their code) can outperform agents in fixed "agentless" harnesses, and there exists a significant gap between single-shot and multi-trial success rates (Pass@1 vs. Pass@5), motivating verifier approaches to our benchmark. Our findings highlight critical challenges in long-horizon code generation, context retrieval, and autonomous experiment execution, establishing AutoExperiment as a new benchmark for evaluating progress in AI-driven scientific experimentation. Our data and code are open-sourced at https://github.com/j1mk1m/AutoExperiment .

Chris Ge, Daria Kryvosheieva, Daniel Fried et al.

As the focus in LLM-based coding shifts from static single-step code generation to multi-step agentic interaction with tools and environments, understanding which tasks will challenge agents and why becomes increasingly difficult. This is compounded by current practice: agent performance is typically measured by aggregate pass rates on benchmarks, but single-number metrics obscure the diversity of tasks within a benchmark. We present a framework for predicting success or failure on individual tasks tailored to the agentic coding regime. Our approach augments Item Response Theory (IRT) with rich features extracted from tasks, including issue statements, repository contexts, solutions, and test cases, and introduces a novel decomposition of agent ability into LLM and scaffold ability components. This parameterization enables us to aggregate evaluation data across heterogeneous leaderboards and accurately predict task-level performance for unseen benchmarks, as well as unseen LLM-scaffold combinations. Our methods have practical utility for benchmark designers, who can better calibrate the difficulty of their new tasks without running computationally expensive agent evaluations.

Atharva Naik, Lawanya Baghel, Dhakshin Govindarajan et al. · cmu

Large Language Models, though successful in code generation, struggle with code quality analysis because they are limited by static training data and can't easily adapt to evolving best practices. We introduce MetaLint, an instruction-following framework that formulates code quality analysis as the task of detecting and fixing problematic semantic code fragments or code idioms based on high-level specifications. Unlike conventional approaches that train models on static code quality conventions, MetaLint employs instruction tuning on synthetic linter-generated data with dynamic conventions to support easy-to-hard generalization, enabling models to adapt to novel or complex code patterns without retraining. To evaluate this, we construct a benchmark of challenging idioms inspired by real-world coding standards such as Python Enhancement Proposals (PEPs) and assess whether MetaLint-trained models reason adaptively or simply memorize. Our results show that MetaLint training improves generalization to unseen idioms. Qwen3-4B attains a 70.37% F-score on a manually curated and challenging PEP idiom detection benchmark, achieving the highest recall (70.43%) among all evaluated models. For localization, it reaches 26.73%, which is a strong outcome for its 4B parameter size and comparable to larger state-of-the-art models such as o3-mini, highlighting its potential for future-proof code quality analysis. Furthermore, MetaLint training enables generalization in idiom detection across model families, model scales, synthetic data from diverse linters, and Java idioms, demonstrating the general applicability of our approach. We plan to release our code and data to enable reproducibility and further work.

Chen Henry Wu, Rishi Shah, Jing Yu Koh et al.

As language models (LMs) are used to build autonomous agents in real environments, ensuring their adversarial robustness becomes a critical challenge. Unlike chatbots, agents are compound systems with multiple components taking actions, which existing LMs safety evaluations do not adequately address. To bridge this gap, we manually create 200 targeted adversarial tasks and evaluation scripts in a realistic threat model on top of VisualWebArena, a real environment for web agents. To systematically examine the robustness of agents, we propose the Agent Robustness Evaluation (ARE) framework. ARE views the agent as a graph showing the flow of intermediate outputs between components and decomposes robustness as the flow of adversarial information on the graph. We find that we can successfully break latest agents that use black-box frontier LMs, including those that perform reflection and tree search. With imperceptible perturbations to a single image (less than 5% of total web page pixels), an attacker can hijack these agents to execute targeted adversarial goals with success rates up to 67%. We also use ARE to rigorously evaluate how the robustness changes as new components are added. We find that inference-time compute that typically improves benign performance can open up new vulnerabilities and harm robustness. An attacker can compromise the evaluator used by the reflexion agent and the value function of the tree search agent, which increases the attack success relatively by 15% and 20%. Our data and code for attacks, defenses, and evaluation are at https://github.com/ChenWu98/agent-attack

Jing Yu Koh, Robert Lo, Lawrence Jang et al.

Autonomous agents capable of planning, reasoning, and executing actions on the web offer a promising avenue for automating computer tasks. However, the majority of existing benchmarks primarily focus on text-based agents, neglecting many natural tasks that require visual information to effectively solve. Given that most computer interfaces cater to human perception, visual information often augments textual data in ways that text-only models struggle to harness effectively. To bridge this gap, we introduce VisualWebArena, a benchmark designed to assess the performance of multimodal web agents on realistic \textit{visually grounded tasks}. VisualWebArena comprises of a set of diverse and complex web-based tasks that evaluate various capabilities of autonomous multimodal agents. To perform on this benchmark, agents need to accurately process image-text inputs, interpret natural language instructions, and execute actions on websites to accomplish user-defined objectives. We conduct an extensive evaluation of state-of-the-art LLM-based autonomous agents, including several multimodal models. Through extensive quantitative and qualitative analysis, we identify several limitations of text-only LLM agents, and reveal gaps in the capabilities of state-of-the-art multimodal language agents. VisualWebArena provides a framework for evaluating multimodal autonomous language agents, and offers insights towards building stronger autonomous agents for the web. Our code, baseline models, and data is publicly available at https://jykoh.com/vwa.

Jacob Mitchell Springer, Suhas Kotha, Daniel Fried et al. · cmu

Bidirectional models are considered essential for strong text embeddings. Recent approaches to adapt autoregressive language models (LMs) into strong text embedding models have largely had the requirement to modify the LM architecture to be bidirectional. We challenge this premise by introducing "echo embeddings" which converts autoregressive LMs into high quality text embedding models without changing the architecture or requiring fine-tuning. By repeating the input and extracting embeddings from the repeated tokens -- which have access to all original tokens -- echo embeddings improve over classical LM embeddings by over 5% in zero-shot settings. Our zero-shot embeddings nearly match those obtained by bidirectionally-converted LMs that undergo additional masked-language modeling training. Echo embeddings are also compatible with supervised fine-tuning, matching or outperforming bidirectionally-converted LMs in an apples-to-apples comparison, even with an identical compute budget during training and inference. Overall, repetition is a simple and effective strategy to circumvent the need for bidirectional attention in embedding models, paving the way towards a unified architecture for all NLP tasks.

Zhiruo Wang, Zhoujun Cheng, Hao Zhu et al.

Language models (LMs) are powerful yet mostly for text generation tasks. Tools have substantially enhanced their performance for tasks that require complex skills. However, many works adopt the term "tool" in different ways, raising the question: What is a tool anyway? Subsequently, where and how do tools help LMs? In this survey, we provide a unified definition of tools as external programs used by LMs, and perform a systematic review of LM tooling scenarios and approaches. Grounded on this review, we empirically study the efficiency of various tooling methods by measuring their required compute and performance gains on various benchmarks, and highlight some challenges and potential future research in the field.

Zhiruo Wang, Daniel Fried, Graham Neubig

Language models (LMs) can solve tasks such as answering questions about tables or images by writing programs. However, using primitive functions often leads to verbose and error-prone programs, and higher-level functions require expert design. To enable better solutions without human labor, we ask code LMs to curate reusable high-level functions, and use them to write solutions. We present TROVE, a training-free method of inducing a verifiable and efficient toolbox of functions, by generating via using, growing, and periodically trimming the toolbox. On 11 datasets from math, table question answering, and image reasoning tasks, TROVE consistently yields simpler solutions with higher accuracy than baselines using CODELLAMA and previous methods using GPT, while using 79-98% smaller toolboxes. TROVE further enables 31% faster and 13% more accurate human verification than baselines. With the same pipeline, it creates diverse functions for varied tasks and datasets, providing insights into their individual characteristics.

Jiaxin Ge, Zora Zhiruo Wang, Xuhui Zhou et al. · allen-ai, cmu

Designing structured visuals such as presentation slides is essential for communicative needs, necessitating both content creation and visual planning skills. In this work, we tackle the challenge of automated slide generation, where models produce slide presentations from natural language (NL) instructions. We first introduce the SlidesBench benchmark, the first benchmark for slide generation with 7k training and 585 testing examples derived from 310 slide decks across 10 domains. SlidesBench supports evaluations that are (i)reference-based to measure similarity to a target slide, and (ii)reference-free to measure the design quality of generated slides alone. We benchmark end-to-end image generation and program generation methods with a variety of models, and find that programmatic methods produce higher-quality slides in user-interactable formats. Built on the success of program generation, we create AutoPresent, an 8B Llama-based model trained on 7k pairs of instructions paired with code for slide generation, and achieve results comparable to the closed-source model GPT-4o. We further explore iterative design refinement where the model is tasked to self-refine its own output, and we found that this process improves the slide's quality. We hope that our work will provide a basis for future work on generating structured visuals.

Zora Zhiruo Wang, Apurva Gandhi, Graham Neubig et al.

To succeed in common digital tasks such as web navigation, agents must carry out a variety of specialized tasks such as searching for products or planning a travel route. To tackle these tasks, agents can bootstrap themselves by learning task-specific skills online through interaction with the web environment. In this work, we demonstrate that programs are an effective representation for skills. We propose agent skill induction (ASI), which allows agents to adapt themselves by inducing, verifying, and utilizing program-based skills on the fly. We start with an evaluation on the WebArena agent benchmark and show that ASI outperforms the static baseline agent and its text-skill counterpart by 23.5% and 11.3% in success rate, mainly thanks to the programmatic verification guarantee during the induction phase. ASI also improves efficiency by reducing 10.7-15.3% of the steps over baselines, by composing primitive actions (e.g., click) into higher-level skills (e.g., search product). We then highlight the efficacy of ASI in remaining efficient and accurate under scaled-up web activities. Finally, we examine the generalizability of induced skills when transferring between websites, and find that ASI can effectively reuse common skills, while also updating incompatible skills to versatile website changes.

Yiqing Xie, Wenxuan Zhou, Pradyot Prakash et al.

Factuality evaluation aims to detect factual errors produced by language models (LMs) and hence guide the development of more factual models. Towards this goal, we train a factuality evaluator, FenCE, that provides LM generators with claim-level factuality feedback. We conduct data augmentation on a combination of public judgment datasets to train FenCE to (1) generate textual critiques along with scores and (2) make claim-level judgment based on diverse source documents obtained by various tools. We then present a framework that leverages FenCE to improve the factuality of LM generators by constructing training data. Specifically, we generate a set of candidate responses, leverage FenCE to revise and score each response without introducing lesser-known facts, and train the generator by preferring highly scored revised responses. Experiments show that our data augmentation methods improve the evaluator's accuracy by 2.9% on LLM-AggreFact. With FenCE, we improve Llama2-7B-chat and Llama3-8B-chat's factuality rate by 16.86% and 14.45% on FActScore, outperforming state-of-the-art factuality finetuning methods by 8.83% and 6.96%.

Yiqing Xie, Alex Xie, Divyanshu Sheth et al.

We present RepoST, a scalable method to construct environments that provide execution feedback for repository-level code generation for both training and evaluation. Unlike existing works that aim to build entire repositories for execution, which is challenging for both human and LLMs, we provide execution feedback with sandbox testing, which isolates a given target function and its dependencies to a separate script for testing. Sandbox testing reduces the complexity of external dependencies and enables constructing environments at a large scale. We use our method to construct RepoST-Train, a large-scale train set with 7,415 functions from 832 repositories. Training with the execution feedback provided by RepoST-Train leads to a performance gain of 5.5% Pass@1 on HumanEval and 3.5% Pass@1 on RepoEval. We also build an evaluation dataset, RepoST-Eval, and benchmark 12 code generation models.

Shenghui Chen, Daniel Fried, Ufuk Topcu

Developing autonomous agents that can strategize and cooperate with humans under information asymmetry is challenging without effective communication in natural language. We introduce a shared-control game, where two players collectively control a token in alternating turns to achieve a common objective under incomplete information. We formulate a policy synthesis problem for an autonomous agent in this game with a human as the other player. To solve this problem, we propose a communication-based approach comprising a language module and a planning module. The language module translates natural language messages into and from a finite set of flags, a compact representation defined to capture player intents. The planning module leverages these flags to compute a policy using an asymmetric information-set Monte Carlo tree search with flag exchange algorithm we present. We evaluate the effectiveness of this approach in a testbed based on Gnomes at Night, a search-and-find maze board game. Results of human subject experiments show that communication narrows the information gap between players and enhances human-agent cooperation efficiency with fewer turns.

Abhishek N. Kulkarni, Andy Liu, Jean-Raphael Gaglione et al. · cmu

In strategic multi-agent sequential interactions, detecting dynamic coalition structures is crucial for understanding how self-interested agents coordinate to influence outcomes. However, natural-language-based interactions introduce unique challenges to coalition detection due to ambiguity over intents and difficulty in modeling players' subjective perspectives. We propose a new method that leverages recent advancements in large language models and game theory to predict dynamic multilateral coalition formation in Diplomacy, a strategic multi-agent game where agents negotiate coalitions using natural language. The method consists of two stages. The first stage extracts the set of agreements discussed by two agents in their private dialogue, by combining a parsing-based filtering function with a fine-tuned language model trained to predict player intents. In the second stage, we define a new metric using the concept of subjective rationalizability from hypergame theory to evaluate the expected value of an agreement for each player. We then compute this metric for each agreement identified in the first stage by assessing the strategic value of the agreement for both players and taking into account the subjective belief of one player that the second player would honor the agreement. We demonstrate that our method effectively detects potential coalition structures in online Diplomacy gameplay by assigning high values to agreements likely to be honored and low values to those likely to be violated. The proposed method provides foundational insights into coalition formation in multi-agent environments with language-based negotiation and offers key directions for future research on the analysis of complex natural language-based interactions between agents.