Isabelle Lee

Ziyi Liu, Soumya Sanyal, Isabelle Lee et al. · amazon-science, uw

In a plethora of recent work, large language models (LLMs) demonstrated impressive reasoning ability, but many proposed downstream reasoning tasks only focus on final answers. Two fundamental questions persist: 1) how consistent is the reasoning, and 2) can models detect unreliable reasoning? In this paper, we investigate self-contradictory (Self-Contra) reasoning, where the model reasoning does not support its answers. To answer 1), we define and assess the Self-Contra rate across three datasets and delve into finer-grained categories of Self-Contra reasoning. We find that LLMs often contradict themselves in reasoning tasks involving contextual information understanding or commonsense. The model may generate correct answers by taking shortcuts in reasoning or overlooking contextual evidence, leading to compromised reasoning. For 2), we task the state-of-the-art model GPT-4 with identifying Self-Contra reasoning and finer-grained fallacies. We find that finer-grained categories enhanced detection can improve GPT-4's ability to detect Self-Contra. However, it is only able to detect Self-Contra with a 52.2% F1 score, much lower compared to 66.7% for humans. Our results indicate that current LLMs lack the robustness necessary for reliable reasoning and we emphasize the urgent need for establishing best practices in comprehensive reasoning evaluations beyond pure performance-based metrics.

Ting-Rui Chiang, Xinyan Velocity Yu, Joshua Robinson et al. · amazon-science, uw

Augmenting a language model (LM) with $k$-nearest neighbors ($k$NN) retrieval on its training data alone can decrease its perplexity, though the underlying reasons for this remain elusive. In this work, we rule out one previously posited possibility -- the "softmax bottleneck." We then create a new dataset to evaluate LM generalization ability in the setting where training data contains additional information that is not causally relevant. This task is challenging even for GPT-3.5 Turbo. We show that, for both GPT-2 and Mistral 7B, $k$NN retrieval augmentation consistently improves performance in this setting. Finally, to make $k$NN retrieval more accessible, we propose using a multi-layer perceptron model that maps datastore keys to values as a drop-in replacement for traditional retrieval. This reduces storage costs by over 25x.

Isabelle Lee, Emmy Liu, Cathy Jiao et al.

Current machine learning models are evaluated through behavioral snapshots, with benchmark accuracies, win rates and outcome-based metrics. Model explanations and evaluations, however, are fundamentally intertwined: understanding why a model produces a behavior can be as important as measuring what it produces. If we trusted interpretability, we argue that it can serve not merely as diagnostics but as a richer and more principled form of model evaluation beyond surface-level performance metrics. We explore three ways interpretability can function evaluatively: (1) fixing problems by identifying the root causes of unwanted behavior, (2) detecting subtly faulty mechanisms that invalidate model outputs, and (3) predicting potential issues before they arise by fully understanding the model's weaknesses. To fulfill its evaluative potential, we argue that interpretability methods must generate claims that are falsifiable, reproducible, and predictive -- that is, interpretability must meet scientific standards.

Hadas Orgad, Fazl Barez, Tal Haklay et al.

Interpretability aims to explain the behavior of deep neural networks. Despite rapid growth, there is mounting concern that much of this work has not translated into practical impact, raising questions about its relevance and utility. This position paper argues that the central missing ingredient is not new methods, but evaluation criteria: interpretability should be evaluated by actionability--the extent to which insights enable concrete decisions and interventions beyond interpretability research itself. We define actionable interpretability along two dimensions--concreteness and validation--and analyze the barriers currently preventing real-world impact. To address these barriers, we identify five domains where interpretability offers unique leverage and present a framework for actionable interpretability with evaluation criteria aligned with practical outcomes. Our goal is not to downplay exploratory research, but to establish actionability as a core objective of interpretability research.

Hannah Murray, Brian Hyeongseok Kim, Isabelle Lee et al. · amazon-science, uw

Large Language Models (LLMs) are becoming ubiquitous, promising automation even in high-stakes scenarios. However, existing evaluation methods often fall short -- benchmarks saturate, accuracy-based metrics are overly simplistic, and many inherently ambiguous problems lack a clear ground truth. Given these limitations, evaluating fairness becomes complex. To address this, we reframe fairness evaluation using Borda scores, a method from voting theory, as a nuanced yet interpretable metric for measuring fairness. Using organ allocation as a case study, we introduce two tasks: (1) Choose-One and (2) Rank-All. In Choose-One, LLMs select a single candidate for a kidney, and we assess fairness across demographics using proportional parity. In Rank-All, LLMs rank all candidates for a kidney, reflecting real-world allocation processes. Since traditional fairness metrics do not account for ranking, we propose a novel application of Borda scoring to capture biases. Our findings highlight the potential of voting-based metrics to provide a richer, more multifaceted evaluation of LLM fairness.

Emmy Liu, Kaiser Sun, Millicent Li et al.

Large language models (LLMs) can perform remarkably complex tasks, yet the fine-grained details of how these capabilities emerge during pretraining remain poorly understood. Scaling laws on validation loss tell us how much a model improves with additional compute, but not what skills it acquires in which order. To remedy this, we propose the Implicit Curriculum Hypothesis: pretraining follows a compositional and predictable curriculum across models and data mixtures. We test this by designing a suite of simple, composable tasks spanning retrieval, morphological transformations, coreference, logical reasoning, and mathematics. Using these tasks, we track emergence points across four model families spanning sizes from 410M-13B parameters. We find that emergence orderings of when models reach fixed accuracy thresholds are strikingly consistent ($ρ= .81$ across 45 model pairs), and that composite tasks most often emerge after their component tasks. Furthermore, we find that this structure is encoded in model representations: tasks with similar function vector representations also tend to follow similar trajectories in training. By using the space of representations derived from our task set, we can effectively predict the training trajectories of simple held-out compositional tasks throughout the course of pretraining ($R^2 = .68$-$.84$ across models) without previously evaluating them. Together, these results suggest that pretraining is more structured than loss curves reveal: skills emerge in a compositional order that is consistent across models and readable from their internals.

Isabelle Lee, Sarah Liaw, Dani Yogatama · amazon-science, uw

Transformer-based large language models (LLMs) have demonstrated remarkable reasoning capabilities such as coding and solving mathematical problems to commonsense inference. While these tasks vary in complexity, they all require models to integrate and compute over structured information. Despite recent efforts to reverse-engineer LLM behavior through controlled experiments, our understanding of how these models internalize and execute complex algorithms remains limited. Progress has largely been confined to small-scale studies or shallow tasks such as basic arithmetic and grammatical pattern matching. One barrier to deeper understanding is the nature of pretraining data -- vast, heterogeneous, and often poorly annotated, making it difficult to isolate mechanisms of reasoning. To bridge this gap, we introduce a large-scale, fully open, complexity-annotated dataset of first-order logic reasoning traces, designed to probe and analyze algorithmic reasoning in LLMs. The dataset consists of 3.5 billion tokens, including 8.8 million LLM-augmented, human-annotated examples and 7.5 million synthetically generated examples. Each synthetic example is verifiably correct, produced by a custom automated theorem solver, and accompanied by metadata tracing its algorithmic provenance. We aim to provide a scalable, interpretable artifact for studying how LLMs learn and generalize symbolic reasoning processes, paving the way for more transparent and targeted investigations into the algorithmic capabilities of modern models.

Isabelle Lee, Joshua Lum, Ziyi Liu et al.

While interpretability research has shed light on some internal algorithms utilized by transformer-based LLMs, reasoning in natural language, with its deep contextuality and ambiguity, defies easy categorization. As a result, formulating clear and motivating questions for circuit analysis that rely on well-defined in-domain and out-of-domain examples required for causal interventions is challenging. Although significant work has investigated circuits for specific tasks, such as indirect object identification (IOI), deciphering natural language reasoning through circuits remains difficult due to its inherent complexity. In this work, we take initial steps to characterize causal reasoning in LLMs by analyzing clear-cut cause-and-effect sentences like "I opened an umbrella because it started raining," where causal interventions may be possible through carefully crafted scenarios using GPT-2 small. Our findings indicate that causal syntax is localized within the first 2-3 layers, while certain heads in later layers exhibit heightened sensitivity to nonsensical variations of causal sentences. This suggests that models may infer reasoning by (1) detecting syntactic cues and (2) isolating distinct heads in the final layers that focus on semantic relationships.

Laiba Mehnaz, Debanjan Mahata, Rakesh Gosangi et al.

Code-switching is the communication phenomenon where speakers switch between different languages during a conversation. With the widespread adoption of conversational agents and chat platforms, code-switching has become an integral part of written conversations in many multi-lingual communities worldwide. This makes it essential to develop techniques for summarizing and understanding these conversations. Towards this objective, we introduce abstractive summarization of Hindi-English code-switched conversations and develop the first code-switched conversation summarization dataset - GupShup, which contains over 6,831 conversations in Hindi-English and their corresponding human-annotated summaries in English and Hindi-English. We present a detailed account of the entire data collection and annotation processes. We analyze the dataset using various code-switching statistics. We train state-of-the-art abstractive summarization models and report their performances using both automated metrics and human evaluation. Our results show that multi-lingual mBART and multi-view seq2seq models obtain the best performances on the new dataset