Roma Patel

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.

Rafael Rodriguez-Sanchez, Benjamin A. Spiegel, Jennifer Wang et al.

We introduce RLang, a domain-specific language (DSL) for communicating domain knowledge to an RL agent. Unlike existing RL DSLs that ground to \textit{single} elements of a decision-making formalism (e.g., the reward function or policy), RLang can specify information about every element of a Markov decision process. We define precise syntax and grounding semantics for RLang, and provide a parser that grounds RLang programs to an algorithm-agnostic \textit{partial} world model and policy that can be exploited by an RL agent. We provide a series of example RLang programs demonstrating how different RL methods can exploit the resulting knowledge, encompassing model-free and model-based tabular algorithms, policy gradient and value-based methods, hierarchical approaches, and deep methods.

Gheorghe Comanici, Eric Bieber, Mike Schaekermann et al. · amazon-science, baidu

In this report, we introduce the Gemini 2.X model family: Gemini 2.5 Pro and Gemini 2.5 Flash, as well as our earlier Gemini 2.0 Flash and Flash-Lite models. Gemini 2.5 Pro is our most capable model yet, achieving SoTA performance on frontier coding and reasoning benchmarks. In addition to its incredible coding and reasoning skills, Gemini 2.5 Pro is a thinking model that excels at multimodal understanding and it is now able to process up to 3 hours of video content. Its unique combination of long context, multimodal and reasoning capabilities can be combined to unlock new agentic workflows. Gemini 2.5 Flash provides excellent reasoning abilities at a fraction of the compute and latency requirements and Gemini 2.0 Flash and Flash-Lite provide high performance at low latency and cost. Taken together, the Gemini 2.X model generation spans the full Pareto frontier of model capability vs cost, allowing users to explore the boundaries of what is possible with complex agentic problem solving.

Ruben Laukkonen, Seb Krier, Chloé Bakalar et al.

Existing alignment research is dominated by concerns about safety and preventing harm: safeguards, controllability, and compliance. This paradigm of alignment parallels early psychology's focus on mental illness: necessary but incomplete. What we call Positive Alignment is the development of AI systems that (i) actively support human and ecological flourishing in a pluralistic, polycentric, context-sensitive, and user-authored way while (ii) remaining safe and cooperative. It is a distinct and necessary agenda within AI alignment research. We argue that several existing failures of alignment (e.g., engagement hacking, loss of human autonomy, failures in truth-seeking, low epistemic humility, error correction, lack of diverse viewpoints, and being primarily reactive rather than proactive) may be better addressed through positive alignment, including cultivating virtues and maximizing human flourishing. We highlight a range of challenges, open questions, and technical directions (e.g., data filtering and upsampling, pre- and post-training, evaluations, collaborative value collection) for different phases of the LLM and agents lifecycle. We end with design principles for promoting disagreement and decentralization through contextual grounding, community customization, continual adaptation, and polycentric governance; that is, many legitimate centers of oversight rather than one institutional or moral chokepoint.

Taylor Sorensen, Pushkar Mishra, Roma Patel et al. · uw

Modelling human variation in rating tasks is crucial for personalization, pluralistic model alignment, and computational social science. We propose representing individuals using natural language value profiles -- descriptions of underlying values compressed from in-context demonstrations -- along with a steerable decoder model that estimates individual ratings from a rater representation. To measure the predictive information in a rater representation, we introduce an information-theoretic methodology and find that demonstrations contain the most information, followed by value profiles, then demographics. However, value profiles effectively compress the useful information from demonstrations (>70% information preservation) and offer advantages in terms of scrutability, interpretability, and steerability. Furthermore, clustering value profiles to identify similarly behaving individuals better explains rater variation than the most predictive demographic groupings. Going beyond test set performance, we show that the decoder predictions change in line with semantic profile differences, are well-calibrated, and can help explain instance-level disagreement by simulating an annotator population. These results demonstrate that value profiles offer novel, predictive ways to describe individual variation beyond demographics or group information.

Charvi Rastogi, Tian Huey Teh, Pushkar Mishra et al.

AI systems crucially rely on human ratings, but these ratings are often aggregated, obscuring the inherent diversity of perspectives in real-world phenomenon. This is particularly concerning when evaluating the safety of generative AI, where perceptions and associated harms can vary significantly across socio-cultural contexts. While recent research has studied the impact of demographic differences on annotating text, there is limited understanding of how these subjective variations affect multimodal safety in generative AI. To address this, we conduct a large-scale study employing highly-parallel safety ratings of about 1000 text-to-image (T2I) generations from a demographically diverse rater pool of 630 raters balanced across 30 intersectional groups across age, gender, and ethnicity. Our study shows that (1) there are significant differences across demographic groups (including intersectional groups) on how severe they assess the harm to be, and that these differences vary across different types of safety violations, (2) the diverse rater pool captures annotation patterns that are substantially different from expert raters trained on specific set of safety policies, and (3) the differences we observe in T2I safety are distinct from previously documented group level differences in text-based safety tasks. To further understand these varying perspectives, we conduct a qualitative analysis of the open-ended explanations provided by raters. This analysis reveals core differences into the reasons why different groups perceive harms in T2I generations. Our findings underscore the critical need for incorporating diverse perspectives into safety evaluation of generative AI ensuring these systems are truly inclusive and reflect the values of all users.

Charvi Rastogi, Tian Huey Teh, Pushkar Mishra et al.

Current text-to-image (T2I) models often fail to account for diverse human experiences, leading to misaligned systems. We advocate for pluralistic alignment, where an AI understands and is steerable towards diverse, and often conflicting, human values. Our work provides three core contributions to achieve this in T2I models. First, we introduce a novel dataset for Diverse Intersectional Visual Evaluation (DIVE) -- the first multimodal dataset for pluralistic alignment. It enable deep alignment to diverse safety perspectives through a large pool of demographically intersectional human raters who provided extensive feedback across 1000 prompts, with high replication, capturing nuanced safety perceptions. Second, we empirically confirm demographics as a crucial proxy for diverse viewpoints in this domain, revealing significant, context-dependent differences in harm perception that diverge from conventional evaluations. Finally, we discuss implications for building aligned T2I models, including efficient data collection strategies, LLM judgment capabilities, and model steerability towards diverse perspectives. This research offers foundational tools for more equitable and aligned T2I systems. Content Warning: The paper includes sensitive content that may be harmful.

Rachel Ma, Lyndon Lam, Benjamin A. Spiegel et al.

It is imperative that robots can understand natural language commands issued by humans. Such commands typically contain verbs that signify what action should be performed on a given object and that are applicable to many objects. We propose a method for generalizing manipulation skills to novel objects using verbs. Our method learns a probabilistic classifier that determines whether a given object trajectory can be described by a specific verb. We show that this classifier accurately generalizes to novel object categories with an average accuracy of 76.69% across 13 object categories and 14 verbs. We then perform policy search over the object kinematics to find an object trajectory that maximizes classifier prediction for a given verb. Our method allows a robot to generate a trajectory for a novel object based on a verb, which can then be used as input to a motion planner. We show that our model can generate trajectories that are usable for executing five verb commands applied to novel instances of two different object categories on a real robot.

Roberto-Rafael Maura-Rivero, Chirag Nagpal, Roma Patel et al.

Current methods that train large language models (LLMs) with reinforcement learning feedback, often resort to averaging outputs of multiple rewards functions during training. This overlooks crucial aspects of individual reward dimensions and inter-reward dependencies that can lead to sub-optimal outcomes in generations. In this work, we show how linear aggregation of rewards exhibits some vulnerabilities that can lead to undesired properties of generated text. We then propose a transformation of reward functions inspired by economic theory of utility functions (specifically Inada conditions), that enhances sensitivity to low reward values while diminishing sensitivity to already high values. We compare our approach to the existing baseline methods that linearly aggregate rewards and show how the Inada-inspired reward feedback is superior to traditional weighted averaging. We quantitatively and qualitatively analyse the difference in the methods, and see that models trained with Inada-transformations score as more helpful while being less harmful.

Ding Wang, Mark Díaz, Charvi Rastogi et al.

Understanding what constitutes safety in AI-generated content is complex. While developers often rely on predefined taxonomies, real-world safety judgments also involve personal, social, and cultural perceptions of harm. This paper examines how annotators evaluate the safety of AI-generated images, focusing on the qualitative reasoning behind their judgments. Analyzing 5,372 open-ended comments, we find that annotators consistently invoke moral, emotional, and contextual reasoning that extends beyond structured safety categories. Many reflect on potential harm to others more than to themselves, grounding their judgments in lived experience, collective risk, and sociocultural awareness. Beyond individual perceptions, we also find that the structure of the task itself -- including annotation guidelines -- shapes how annotators interpret and express harm. Guidelines influence not only which images are flagged, but also the moral judgment behind the justifications. Annotators frequently cite factors such as image quality, visual distortion, and mismatches between prompt and output as contributing to perceived harm dimensions, which are often overlooked in standard evaluation frameworks. Our findings reveal that existing safety pipelines miss critical forms of reasoning that annotators bring to the task. We argue for evaluation designs that scaffold moral reflection, differentiate types of harm, and make space for subjective, context-sensitive interpretations of AI-generated content.

Michael A. Lepori, Jennifer Hu, Ishita Dasgupta et al.

Language models (LMs) are used for a diverse range of tasks, from question answering to writing fantastical stories. In order to reliably accomplish these tasks, LMs must be able to discern the modal category of a sentence (i.e., whether it describes something that is possible, impossible, completely nonsensical, etc.). However, recent studies have called into question the ability of LMs to categorize sentences according to modality (Michaelov et al., 2025; Kauf et al., 2023). In this work, we identify linear representations that discriminate between modal categories within a variety of LMs, or modal difference vectors. Analysis of modal difference vectors reveals that LMs have access to more reliable modal categorization judgments than previously reported. Furthermore, we find that modal difference vectors emerge in a consistent order as models become more competent (i.e., through training steps, layers, and parameter count). Notably, we find that modal difference vectors identified within LM activations can be used to model fine-grained human categorization behavior. This potentially provides a novel view into how human participants distinguish between modal categories, which we explore by correlating projections along modal difference vectors with human participants' ratings of interpretable features. In summary, we derive new insights into LM modal categorization using techniques from mechanistic interpretability, with the potential to inform our understanding of modal categorization in humans.

Ian Gemp, Roma Patel, Yoram Bachrach et al.

Mathematical models of interactions among rational agents have long been studied in game theory. However these interactions are often over a small set of discrete game actions which is very different from how humans communicate in natural language. To bridge this gap, we introduce a framework that allows equilibrium solvers to work over the space of natural language dialogue generated by large language models (LLMs). Specifically, by modelling the players, strategies and payoffs in a "game" of dialogue, we create a binding from natural language interactions to the conventional symbolic logic of game theory. Given this binding, we can ask existing game-theoretic algorithms to provide us with strategic solutions (e.g., what string an LLM should generate to maximize payoff in the face of strategic partners or opponents), giving us predictors of stable, rational conversational strategies. We focus on three domains that require different negotiation strategies: scheduling meetings, trading fruit and debate, and evaluate an LLM's generated language when guided by solvers. We see that LLMs that follow game-theory solvers result in dialogue generations that are less exploitable than the control (no guidance from solvers), and the language generated results in higher rewards, in all negotiation domains. We discuss future implications of this work, and how game-theoretic solvers that can leverage the expressivity of natural language can open up a new avenue of guiding language research.

Eric Hsiung, Hiloni Mehta, Junchi Chu et al.

Recent work on using natural language to specify commands to robots has grounded that language to LTL. However, mapping natural language task specifications to LTL task specifications using language models require probability distributions over finite vocabulary. Existing state-of-the-art methods have extended this finite vocabulary to include unseen terms from the input sequence to improve output generalization. However, novel out-of-vocabulary atomic propositions cannot be generated using these methods. To overcome this, we introduce an intermediate contextual query representation which can be learned from single positive task specification examples, associating a contextual query with an LTL template. We demonstrate that this intermediate representation allows for generalization over unseen object references, assuming accurate groundings are available. We compare our method of mapping natural language task specifications to intermediate contextual queries against state-of-the-art CopyNet models capable of translating natural language to LTL, by evaluating whether correct LTL for manipulation and navigation task specifications can be output, and show that our method outperforms the CopyNet model on unseen object references. We demonstrate that the grounded LTL our method outputs can be used for planning in a simulated OO-MDP environment. Finally, we discuss some common failure modes encountered when translating natural language task specifications to grounded LTL.

Alexander Ku, Peter Anderson, Roma Patel et al.

We introduce Room-Across-Room (RxR), a new Vision-and-Language Navigation (VLN) dataset. RxR is multilingual (English, Hindi, and Telugu) and larger (more paths and instructions) than other VLN datasets. It emphasizes the role of language in VLN by addressing known biases in paths and eliciting more references to visible entities. Furthermore, each word in an instruction is time-aligned to the virtual poses of instruction creators and validators. We establish baseline scores for monolingual and multilingual settings and multitask learning when including Room-to-Room annotations. We also provide results for a model that learns from synchronized pose traces by focusing only on portions of the panorama attended to in human demonstrations. The size, scope and detail of RxR dramatically expands the frontier for research on embodied language agents in simulated, photo-realistic environments.

Thao Nguyen, Nakul Gopalan, Roma Patel et al.

Natural language object retrieval is a highly useful yet challenging task for robots in human-centric environments. Previous work has primarily focused on commands specifying the desired object's type such as "scissors" and/or visual attributes such as "red," thus limiting the robot to only known object classes. We develop a model to retrieve objects based on descriptions of their usage. The model takes in a language command containing a verb, for example "Hand me something to cut," and RGB images of candidate objects and selects the object that best satisfies the task specified by the verb. Our model directly predicts an object's appearance from the object's use specified by a verb phrase. We do not need to explicitly specify an object's class label. Our approach allows us to predict high level concepts like an object's utility based on the language query. Based on contextual information present in the language commands, our model can generalize to unseen object classes and unknown nouns in the commands. Our model correctly selects objects out of sets of five candidates to fulfill natural language commands, and achieves an average accuracy of 62.3% on a held-out test set of unseen ImageNet object classes and 53.0% on unseen object classes and unknown nouns. Our model also achieves an average accuracy of 54.7% on unseen YCB object classes, which have a different image distribution from ImageNet objects. We demonstrate our model on a KUKA LBR iiwa robot arm, enabling the robot to retrieve objects based on natural language descriptions of their usage. We also present a new dataset of 655 verb-object pairs denoting object usage over 50 verbs and 216 object classes.

Yoonseon Oh, Roma Patel, Thao Nguyen et al.

Often times, we specify tasks for a robot using temporal language that can also span different levels of abstraction. The example command ``go to the kitchen before going to the second floor'' contains spatial abstraction, given that ``floor'' consists of individual rooms that can also be referred to in isolation ("kitchen", for example). There is also a temporal ordering of events, defined by the word "before". Previous works have used Linear Temporal Logic (LTL) to interpret temporal language (such as "before"), and Abstract Markov Decision Processes (AMDPs) to interpret hierarchical abstractions (such as "kitchen" and "second floor"), separately. To handle both types of commands at once, we introduce the Abstract Product Markov Decision Process (AP-MDP), a novel approach capable of representing non-Markovian reward functions at different levels of abstractions. The AP-MDP framework translates LTL into its corresponding automata, creates a product Markov Decision Process (MDP) of the LTL specification and the environment MDP, and decomposes the problem into subproblems to enable efficient planning with abstractions. AP-MDP performs faster than a non-hierarchical method of solving LTL problems in over 95% of tasks, and this number only increases as the size of the environment domain increases. We also present a neural sequence-to-sequence model trained to translate language commands into LTL expression, and a new corpus of non-Markovian language commands spanning different levels of abstraction. We test our framework with the collected language commands on a drone, demonstrating that our approach enables a robot to efficiently solve temporal commands at different levels of abstraction.

Najoung Kim, Roma Patel, Adam Poliak et al.

We introduce a set of nine challenge tasks that test for the understanding of function words. These tasks are created by structurally mutating sentences from existing datasets to target the comprehension of specific types of function words (e.g., prepositions, wh-words). Using these probing tasks, we explore the effects of various pretraining objectives for sentence encoders (e.g., language modeling, CCG supertagging and natural language inference (NLI)) on the learned representations. Our results show that pretraining on language modeling performs the best on average across our probing tasks, supporting its widespread use for pretraining state-of-the-art NLP models, and CCG supertagging and NLI pretraining perform comparably. Overall, no pretraining objective dominates across the board, and our function word probing tasks highlight several intuitive differences between pretraining objectives, e.g., that NLI helps the comprehension of negation.

Alex Wang, Jan Hula, Patrick Xia et al.

Natural language understanding has recently seen a surge of progress with the use of sentence encoders like ELMo (Peters et al., 2018a) and BERT (Devlin et al., 2019) which are pretrained on variants of language modeling. We conduct the first large-scale systematic study of candidate pretraining tasks, comparing 19 different tasks both as alternatives and complements to language modeling. Our primary results support the use language modeling, especially when combined with pretraining on additional labeled-data tasks. However, our results are mixed across pretraining tasks and show some concerning trends: In ELMo's pretrain-then-freeze paradigm, random baselines are worryingly strong and results vary strikingly across target tasks. In addition, fine-tuning BERT on an intermediate task often negatively impacts downstream transfer. In a more positive trend, we see modest gains from multitask training, suggesting the development of more sophisticated multitask and transfer learning techniques as an avenue for further research.

Benjamin Nye, Junyi Jessy Li, Roma Patel et al.

We present a corpus of 5,000 richly annotated abstracts of medical articles describing clinical randomized controlled trials. Annotations include demarcations of text spans that describe the Patient population enrolled, the Interventions studied and to what they were Compared, and the Outcomes measured (the `PICO' elements). These spans are further annotated at a more granular level, e.g., individual interventions within them are marked and mapped onto a structured medical vocabulary. We acquired annotations from a diverse set of workers with varying levels of expertise and cost. We describe our data collection process and the corpus itself in detail. We then outline a set of challenging NLP tasks that would aid searching of the medical literature and the practice of evidence-based medicine.

Roma Patel, Yinfei Yang, Iain Marshall et al.

Medical professionals search the published literature by specifying the type of patients, the medical intervention(s) and the outcome measure(s) of interest. In this paper we demonstrate how features encoding syntactic patterns improve the performance of state-of-the-art sequence tagging models (both linear and neural) for information extraction of these medically relevant categories. We present an analysis of the type of patterns exploited, and the semantic space induced for these, i.e., the distributed representations learned for identified multi-token patterns. We show that these learned representations differ substantially from those of the constituent unigrams, suggesting that the patterns capture contextual information that is otherwise lost.