24 accepted papers
Early children's developmental trajectories set up a natural goal for sample-efficient pretraining of vision foundation models. We introduce BabyVLM-V2, a developmentally grounded framework for infant-inspired vision-language modeling that extensively improves upon BabyVLM-V1 through a longitudinal,
Entity tracking (ET), the ability to keep track of states, is a fundamental skill that underlies complex reasoning. An increasing amount of work investigates how transformer language models (LMs) solve entity binding *without* state changes; however, there is limited understanding of how non-toy LMs…
Interpretability research on large language models (LLMs) has produced methods that align model components to high-level concepts, yet their use has been accompanied by recurring failures: findings that do not generalise, and causal language that outruns the evidence. Our position is that Pearl’s ca…
A central aim of interpretability tools applied to language models is to recover meaningful concepts from model activations. Existing feature extraction methods focus on single activations regardless of the context, implicitly assuming independence (and therefore stationarity). This leaves open whet…
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…
Property inheritance—a phenomenon where novel properties are projected from higher level categories (e.g., birds) to lower level ones (e.g., sparrows)—provides a unique window into how humans organize and deploy conceptual knowledge. It is debated whether this ability arises due to explicitly stored…
Autoregressive transformer language models (LMs) possess strong syntactic abilities, often successfully handling phenomena from agreement to NPI licensing. However, the features they use to incrementally process their linguistic input are not well understood. In this paper, we fill this gap by study…
Human bilinguals often use similar brain regions to process multiple languages, depending on when they learned their second language and their proficiency. In large language models (LLMs), how are multiple languages learned and encoded? In this work, we explore the extent to which LLMs share represe…
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…
We introduce NNsight and NDIF, technologies that work in tandem to enable scientific study of the representations and computations learned by very large neural networks. NNsight is an open-source system that extends PyTorch to introduce deferred remote execution. The National Deep Inference Fabric (…
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…
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
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…
We report the presence of a simple neural mechanism that represents an input-output function as a vector within autoregressive transformer language models (LMs). Using causal mediation analysis on a diverse range of in-context-learning (ICL) tasks, we find that a small number attention heads transpo…
In-context learning (ICL) is now a common method for teaching large language models (LLMs) new tasks: given labeled examples in the input context, the LLM learns to perform the task without weight updates. Do models guided via ICL infer the underlying structure of the task defined by the context, or…
Accurate syntactic representations are essential for robust generalization in natural language. Recent work has found that pre-training can teach language models to rely on hierarchical syntactic features—as opposed to incorrect linear features—when performing tasks after fine-tuning. We test what a…
Targeted syntactic evaluations of language models ask whether models show stable preferences for syntactically acceptable content over minimal-pair unacceptable inputs. Our best syntactic evaluation datasets, however, provide substantially less linguistic context than models receive during pretraini…
Large language models show impressive results on few-shot NLP tasks. However, these models are memory and computation-intensive. Meta-training allows one to leverage smaller models for few-shot generalization in a domain-general and task-agnostic manner; however, these methods alone results in model…
We present the results of the NLP Community Metasurvey. Run from May to June 2022, it elicited opinions on controversial issues, including industry influence in the field, concerns about AGI, and ethics. Our results put concrete numbers to several controversies: For example, respondents are split in…
The language of Twitter differs significantly from that of other domains commonly included in large language model training. While tweets are typically multilingual and contain informal language, including emoji and hashtags, most pre-trained language models for Twitter are either monolingual, adapt…
Relations between words are governed by hierarchical structure rather than linear ordering. Sequence-to-sequence (seq2seq) models, despite their success in downstream NLP applications, often fail to generalize in a hierarchy-sensitive manner when performing syntactic transformations—for example, tra…
In text classification tasks, useful information is encoded in the label names. Label semantic aware systems have leveraged this information for improved text classification performance during fine-tuning and prediction. However, use of label-semantics during pre-training has not been extensively ex…
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…
Neural topic models can augment or replace bag-of-words inputs with the learned representations of deep pre-trained transformer-based word prediction models. One added benefit when using representations from multilingual models is that they facilitate zero-shot polylingual topic modeling. However, w…