63 accepted papers
Text-to-Image (T2I) models have advanced rapidly, yet they remain vulnerable to semantic leakage, the unintended transfer of semantically related features between distinct entities. Existing mitigation strategies are often optimization-based or dependent on external inputs. We introduce **DeLeaker**…
Despite the central role of attention heads in Transformers, we lack tools to understand why a model attends to a particular token. To address this, we study the query-key (QK) space -- the bilinear joint embedding space between queries and keys. We present a contrastive covariance method to decompo…
Protein sequences are abundant in repeating segments, both as exact copies and as approximate segments with mutations. These repeats are important for protein structure and function, motivating decades of algorithmic work on repeat identification. Recent work has shown that protein language models (…
How do language models (LMs) represent characters’ beliefs, especially when those beliefs may differ from reality? This question lies at the heart of understanding the Theory of Mind (ToM) capabilities of LMs. We analyze LMs' ability to reason about characters’ beliefs using causal mediation and abs…
As large language models (LLMs) evolve from conversational assistants into autonomous agents, evaluating the safety of their actions becomes critical. Prior safety benchmarks have primarily focused on preventing generation of harmful content, such as toxic text. However, they overlook the challenge…
How does the conversational past of large language models (LLMs) influence their future performance? Recent work suggests that LLMs are affected by their conversational history in unexpected ways. For instance, hallucinations in prior interactions may influence subsequent model responses. In this wo…
Multiple-choice question answering (MCQA) is a key competence of performant transformer language models that is tested by mainstream benchmarks. However, recent evidence shows that models can have quite a range of performance, particularly when the task format is diversified slightly (such as by shu…
Do large language models (LLMs) solve reasoning tasks by learning robust generalizable algorithms, or do they memorize training data? To investigate this question, we use arithmetic reasoning as a representative task. Using causal analysis, we identify a subset of the model (a circuit) that explains…
We investigate how large language models (LLMs) perform latent multi-hop reasoning in prompts like “Wolfgang Amadeus Mozart’s mother’s spouse is”. To analyze this process, we introduce logit flow, an interpretability method that traces how logits propagate across layers and positions toward the fina
Compositionality is a cognitive mechanism that allows humans to systematically combine known concepts in novel ways. This study demonstrates how artificial neural agents acquire and utilize compositional generalization to describe previously unseen images. Our method, termed \`\`Composition through…
We present Jamba, a novel hybrid Transformer-Mamba mixture-of-experts (MoE) architecture. Jamba interleaves blocks of Transformer and Mamba layers, enjoying the benefits of both model families. MoE is added in some of these layers to increase model capacity while keeping active parameter usage manag…
Large language models (LLMs) often produce errors, including factual inaccuracies, biases, and reasoning failures, collectively referred to as "hallucinations". Recent studies have demonstrated that LLMs' internal states encode information regarding the truthfulness of their outputs, and that this i…
How can we know whether new mechanistic interpretability methods achieve real improvements? In pursuit of lasting evaluation standards, we propose MIB, a Mechanistic Interpretability Benchmark, with two tracks spanning four tasks and five models. MIB favors methods that precisely and concisely recov…
When prompted to think step-by-step, language models (LMs) produce a chain of thought (CoT), a sequence of reasoning steps that the model supposedly used to produce its prediction. Despite much work on CoT prompting, it is unclear if reasoning verbalized in a CoT is faithful to the models’ parametri
Text-to-image (T2I) diffusion models rely on encoded prompts to guide the image generation process. Typically, these prompts are extended to a fixed length by appending padding tokens to the input. Despite being a default practice, the influence of padding tokens on the image generation process has…
A widely used strategy to discover and understand language model mechanisms is circuit analysis. A circuit is a minimal subgraph of a model’s computation graph that executes a specific task. We identify a gap in existing circuit discovery methods: they assume circuits are position-invariant, treatin…
Language models (LMs) risk inadvertently memorizing and divulging sensitive or personally identifiable information (PII) seen in training data, causing privacy concerns. Current approaches to address this issue involve costly dataset scrubbing, or model filtering through unlearning and model editing…
Sparse Autoencoders (SAEs) have been proposed as an unsupervised approach to learn a decomposition of a model’s latent space. This enables useful applications, such as fine-grained steering of model outputs without requiring labeled data. Current steering methods identify SAE features to target by a
Vision-Language models (VLMs) show impressive abilities to answer questions on visual inputs (e.g., counting objects in an image), yet demonstrate higher accuracies when performing an analogous task on text (e.g., counting words in a text). We investigate this accuracy gap by identifying and compari…
We introduce methods for discovering and applying **sparse feature circuits**. These are causally implicated subnetworks of human-interpretable features for explaining language model behaviors. Circuits identified in prior work consist of polysemantic and difficult-to-interpret units like attention…
Prior work on large language model (LLM) hallucinations has associated them with model uncertainty or inaccurate knowledge. In this work, we define and investigate a distinct type of hallucination, where a model can consistently answer a question correctly, but a seemingly trivial perturbation, whic
Emergent Communication (EC) provides a unique window into the language systems that emerge autonomously when agents are trained to jointly achieve shared goals. However, it is difficult to interpret EC and evaluate its relationship with natural languages (NL). This study employs unsupervised neural…
Local explanation methods highlight the input tokens that have a considerable impact on the outcome of classifying the document at hand. For example, the Anchor algorithm applies a statistical analysis of the sensitivity of the classifier to changes in the token. Aggregating local explanations over…
Understanding how Transformer-based Language Models (LMs) learn and recall information is a key goal of the deep learning community. Recent interpretability methods project weights and hidden states obtained from the forward pass to the models’ vocabularies, helping to uncover how information flows…
Artificial agents that learn to communicate in order to accomplish a given task acquire communication protocols that are typically opaque to a human. A large body of work has attempted to evaluate the emergent communication via various evaluation measures, with **compositionality** featuring as a pr…
Despite their widespread use, the mechanisms by which large language models (LLMs) represent and regulate uncertainty in next-token predictions remain largely unexplored. This study investigates two critical components believed to influence this uncertainty: the recently discovered entropy neurons a…
Recent work has compared neural network representations via similarity-based analyses to improve model interpretation. The quality of a similarity measure is typically evaluated by its success in assigning a high score to representations that are expected to be matched. However, existing similarity…
Text-to-image diffusion models (T2I) use a latent representation of a text prompt to guide the image generation process. However, the process by which the encoder produces the text representation is unknown. We propose the Diffusion Lens, a method for analyzing the text encoder of T2I models by gene…
Parameter efficient finetuning methods like low-rank adaptation (LoRA) aim to reduce the computational costs of finetuning pretrained Language Models (LMs). Enabled by these low-rank settings, we propose an even more efficient optimization strategy: Fast Forward, a simple and effective approach to a…
Fine-tuning on generalized tasks such as instruction following, code generation, and mathematics has been shown to enhance language models' performance on a range of tasks. Nevertheless, explanations of how such fine-tuning influences the internal computations in these models remain elusive. We stud…
Mitigating social biases typically requires identifying the social groups associated with each data sample. In this paper, we present DAFair, a novel approach to address social bias in language models. Unlike traditional methods that rely on explicit demographic labels, our approach does not require…
Much of the knowledge encoded in transformer language models (LMs) may be expressed in terms of relations: relations between words and their synonyms, entities and their attributes, etc. We show that, for a subset of relations, this computation is well-approximated by a single linear transformation…
Our world is marked by unprecedented technological, global, and socio-political transformations, posing a significant challenge to textto-image generative models. These models encode factual associations within their parameters that can quickly become outdated, diminishing their utility for end-user…
When artificial agents are jointly trained to perform collaborative tasks using a communication channel, they develop opaque goal-oriented communication protocols. Good task performance is often considered sufficient evidence that meaningful communication is taking place, but existing empirical resu…
Mathematical reasoning in large language models (LMs) has garnered significant attention in recent work, but there is a limited understanding of how these models process and store information related to arithmetic tasks within their architecture. In order to improve our understanding of this aspect…
Models trained on real-world data tend to imitate and amplify social biases. Common methods to mitigate biases require prior information on the types of biases that should be mitigated (e.g., gender or racial bias) and the social groups associated with each data sample. In this work, we introduce BL…
Text-to-image diffusion models often make implicit assumptions about the world when generating images. While some assumptions are useful (e.g., the sky is blue), they can also be outdated, incorrect, or reflective of social biases present in the training data. Thus, there is a need to control thes…
The field of emergent communication aims to understand the characteristics of communication as it emerges from artificial agents solving tasks that require information exchange. Communication with discrete messages is considered a desired characteristic, for scientific and applied reasons. However,…
Recent work has shown exciting promise in updating large language models with new memories, so as to replace obsolete information or add specialized knowledge. However, this line of work is predominantly limited to updating single associations. We develop MEMIT, a method for directly updating a lang…
The sequence alignment problem is one of the most fundamental problems in bioinformatics and a plethora of methods were devised to tackle it. Here we introduce BetaAlign, a methodology for aligning sequences using an NLP approach. BetaAlign accounts for the possible variability of the evolutionary p…
When applied to processing long text, Large Language Models (LLMs) are limited by their context window. Existing efforts to address this limitation involve training specialized architectures, and cannot be easily applied to off- the-shelf LLMs. We present Parallel Context Windows (PCW), a method tha…
Natural language processing models tend to learn and encode social biases present in the data. One popular approach for addressing such biases is to eliminate encoded information from the model’s representations. However, current methods are restricted to removing only linearly encoded information.…
Recent advances in interpretability suggest we can project weights and hidden states of transformer-based language models (LMs) to their vocabulary, a transformation that makes them more human interpretable. In this paper, we investigate LM attention heads and memory values, the vectors the models d…
Dual encoders are now the dominant architecture for dense retrieval. Yet, we have little understanding of how they represent text, and why this leads to good performance. In this work, we shed light on this question via distributions over the vocabulary. We propose to interpret the vector representa…
Animacy—whether an entity is alive and sentient—is fundamental to cognitive processing, impacting areas such as memory, vision, and language. However, animacy is not always expressed directly in language: in English it often manifests indirectly, in the form of selectional constraints on verbs and a…
Most evaluations of attribution methods focus on the English language. In this work, we present a multilingual approach for evaluating attribution methods for the Natural Language Inference (NLI) task in terms of faithfulness and plausibility.First, we introduce a novel cross-lingual strategy to mea…
Common studies of gender bias in NLP focus either on extrinsic bias measured by model performance on a downstream task or on intrinsic bias found in models’ internal representations. However, the relationship between extrinsic and intrinsic bias is relatively unknown. In this work, we illuminate thi…
We analyze the storage and recall of factual associations in autoregressive transformer language models, finding evidence that these associations correspond to localized, directly-editable computations. We first develop a causal intervention for identifying neuron activations that are decisive in a…
Neural networks are known to exploit spurious artifacts (or shortcuts) that co-occur with a target label, exhibiting heuristic memorization. On the other hand, networks have been shown to memorize training examples, resulting in example-level memorization. These kinds of memorization impede generali…
While many studies have shown that linguistic information is encoded in hidden word representations, few have studied individual neurons, to show how and in which neurons it is encoded. Among these, the common approach is to use an external probe to rank neurons according to their relevance to some…
Natural Language Inference (NLI) models are known to learn from biases and artefacts within their training data, impacting how well they generalise to other unseen datasets. Existing de-biasing approaches focus on preventing the models from learning these biases, which can result in restrictive mode…
Targeted syntactic evaluations have demonstrated the ability of language models to perform subject-verb agreement given difficult contexts. To elucidate the mechanisms by which the models accomplish this behavior, this study applies causal mediation analysis to pre-trained neural language models. We…
Model robustness to bias is often determined by the generalization on carefully designed out-of-distribution datasets. Recent debiasing methods in natural language understanding (NLU) improve performance on such datasets by pressuring models into making unbiased predictions. An underlying assumption…
Invariant Risk Minimization (IRM) is a recently proposed framework for out-of-distribution (o.o.d) generalization. Most of the studies on IRM so far have focused on theoretical results, toy problems, and simple models. In this work, we investigate the applicability of IRM to bias mitigation-a speci…
State-of-the-art natural language processing (NLP) models often learn to model dataset biases and surface form correlations instead of features that target the intended underlying task. Previous work has demonstrated effective methods to circumvent these issues when knowledge of the bias is availabl…
Self-supervised speech representation learning has recently been a prosperous research topic. Many algorithms have been proposed for learning useful representations from large-scale unlabeled data, and their applications to a wide range of speech tasks have also been investigated. However, there has…
While large-scale pretrained language models have obtained impressive results when fine-tuned on a wide variety of tasks, they still often suffer from overfitting in low-resource scenarios. Since such models are general-purpose feature extractors, many of these features are inevitably irrelevant for…
The dependency of the generalization error of neural networks on model and dataset size is of critical importance both in practice and for understanding the theory of neural networks. Nevertheless, the functional form of this dependency remains elusive. In this work, we present a functional form whi…
Many interpretation methods for neural models in natural language processing investigate how information is encoded inside hidden representations. However, these methods can only measure whether the information exists, not whether it is actually used by the model. We propose a methodology grounded…
Neural machine translation (NMT) models learn representations containing substantial linguistic information. However, it is not clear if such information is fully distributed or if some of it can be attributed to individual neurons. We develop unsupervised methods for discovering important neurons i…
Character-based neural machine translation (NMT) models alleviate out-of-vocabulary issues, learn morphology, and move us closer to completely end-to-end translation systems. Unfortunately, they are also very brittle and easily falter when presented with noisy data. In this paper, we confront NMT m…
Neural networks have become ubiquitous in automatic speech recognition systems. While neural networks are typically used as acoustic models in more complex systems, recent studies have explored end-to-end speech recognition systems based on neural networks, which can be trained to directly predict t…
There is a lot of research interest in encoding variable length sentences into fixed length vectors, in a way that preserves the sentence meanings. Two common methods include representations based on averaging word vectors, and representations based on the hidden states of recurrent neural networks…