English
Related papers

Related papers: On Linear Representations and Pretraining Data Fre…

200 papers

Pretrained Language Models (LMs) have demonstrated ability to perform numerical reasoning by extrapolating from a few examples in few-shot settings. However, the extent to which this extrapolation relies on robust reasoning is unclear. In…

Computation and Language · Computer Science 2023-03-17 Yasaman Razeghi , Robert L. Logan , Matt Gardner , Sameer Singh

Recent works have argued that high-level semantic concepts are encoded "linearly" in the representation space of large language models. In this work, we study the origins of such linear representations. To that end, we introduce a simple…

Computation and Language · Computer Science 2024-03-07 Yibo Jiang , Goutham Rajendran , Pradeep Ravikumar , Bryon Aragam , Victor Veitch

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 and Language · Computer Science 2024-02-19 Evan Hernandez , Arnab Sen Sharma , Tal Haklay , Kevin Meng , Martin Wattenberg , Jacob Andreas , Yonatan Belinkov , David Bau

As Large Language Models (LLMs) become increasingly widespread, understanding how specific training data shapes their outputs is crucial for transparency, accountability, privacy, and fairness. To explore how LLMs leverage and replicate…

Computation and Language · Computer Science 2025-07-03 Arthur Wuhrmann , Anastasiia Kucherenko , Andrei Kucharavy

Large Language Models (LLMs) are capable of recalling multilingual factual knowledge present in their pretraining data. However, most studies evaluate only the final model, leaving the development of factual recall and crosslingual…

Computation and Language · Computer Science 2025-10-08 Yihong Liu , Mingyang Wang , Amir Hossein Kargaran , Felicia Körner , Ercong Nie , Barbara Plank , François Yvon , Hinrich Schütze

Understanding and mitigating hallucinations in Large Language Models (LLMs) is crucial for ensuring reliable content generation. While previous research has primarily focused on "when" LLMs hallucinate, our work explains "why" and directly…

Computation and Language · Computer Science 2025-07-15 Yuan He , Bailan He , Zifeng Ding , Alisia Lupidi , Yuqicheng Zhu , Shuo Chen , Caiqi Zhang , Jiaoyan Chen , Yunpu Ma , Volker Tresp , Ian Horrocks

Large amounts of training data are one of the major reasons for the high performance of state-of-the-art NLP models. But what exactly in the training data causes a model to make a certain prediction? We seek to answer this question by…

Large language models (LLMs) work by manipulating the geometry of input embedding vectors over multiple layers. Here, we ask: how are the input vocabulary representations of language models structured, and how and when does this structure…

Computation and Language · Computer Science 2025-10-10 Isabel Papadimitriou , Jacob Prince

Large Language Models (LLMs) have impressive capabilities, but are prone to outputting falsehoods. Recent work has developed techniques for inferring whether a LLM is telling the truth by training probes on the LLM's internal activations.…

Artificial Intelligence · Computer Science 2024-08-20 Samuel Marks , Max Tegmark

Language model (LM) pre-training is useful in many language processing tasks. But can pre-trained LMs be further leveraged for more general machine learning problems? We propose an approach for using LMs to scaffold learning and…

Language models (LMs) are pretrained to imitate internet text, including content that would violate human preferences if generated by an LM: falsehoods, offensive comments, personally identifiable information, low-quality or buggy code, and…

Computation and Language · Computer Science 2023-06-16 Tomasz Korbak , Kejian Shi , Angelica Chen , Rasika Bhalerao , Christopher L. Buckley , Jason Phang , Samuel R. Bowman , Ethan Perez

Large language models (LLMs) are trained on vast amounts of text from the internet, which contains both factual and misleading information about the world. While unintuitive from a classic view of LMs, recent work has shown that the truth…

Computation and Language · Computer Science 2024-02-07 Nitish Joshi , Javier Rando , Abulhair Saparov , Najoung Kim , He He

How do language models learn to make predictions during pre-training? To study this, we extract learning curves from five autoregressive English language model pre-training runs, for 1M unseen tokens in context. We observe that the language…

Computation and Language · Computer Science 2024-08-01 Tyler A. Chang , Zhuowen Tu , Benjamin K. Bergen

We investigate how LLMs encode sociodemographic attributes of human conversational partners inferred from indirect cues such as names and occupations. We show that LLMs develop linear representations of user demographics within activation…

Artificial Intelligence · Computer Science 2025-12-12 Paul Bouchaud , Pedro Ramaciotti

To effectively perform the task of next-word prediction, long short-term memory networks (LSTMs) must keep track of many types of information. Some information is directly related to the next word's identity, but some is more secondary…

Computation and Language · Computer Science 2021-06-01 Qingfeng Lan , Luke Kumar , Martha White , Alona Fyshe

How large language models internally represent high-level behaviors is a core interpretability question with direct relevance to AI safety: it determines what we can detect, audit, or intervene on. Recent work has shown that traits such as…

Computation and Language · Computer Science 2026-05-14 Viktor Moskvoretskii , Dominik Glandorf , Jorge Medina Moreira , Tanja Käser , Robert West

Pre-training language models (LMs) on large-scale unlabeled text data makes the model much easier to achieve exceptional downstream performance than their counterparts directly trained on the downstream tasks. In this work, we study what…

Computation and Language · Computer Science 2022-02-21 Cheng-Han Chiang , Hung-yi Lee

We investigate the extent to which modern, neural language models are susceptible to structural priming, the phenomenon whereby the structure of a sentence makes the same structure more probable in a follow-up sentence. We explore how…

Computation and Language · Computer Science 2022-06-30 Arabella Sinclair , Jaap Jumelet , Willem Zuidema , Raquel Fernández

We show that language models' activations linearly encode when information was learned during training. Our setup involves creating a model with a known training order by sequentially fine-tuning Llama-3.2-1B on six disjoint but otherwise…

Machine Learning · Computer Science 2025-09-23 Dmitrii Krasheninnikov , Richard E. Turner , David Krueger

In autoregressive large language models (LLMs), temporal straightening offers an account of how the next-token prediction objective shapes representations. Models learn to progressively straighten the representational trajectory of input…

Artificial Intelligence · Computer Science 2026-04-28 Jack King , Evelina Fedorenko , Eghbal A. Hosseini
‹ Prev 1 2 3 10 Next ›