English
Related papers

Related papers: Monitoring Latent World States in Language Models …

200 papers

This study investigates the behaviors of Large Language Models (LLMs) when faced with conflicting prompts versus their internal memory. This will not only help to understand LLMs' decision mechanism but also benefit real-world applications,…

Computation and Language · Computer Science 2024-02-21 Jiahao Ying , Yixin Cao , Kai Xiong , Yidong He , Long Cui , Yongbin Liu

Large Language Models (LLMs) are often used as automated judges to evaluate text, but their effectiveness can be hindered by various unintentional biases. We propose using linear classifying probes, trained by leveraging differences between…

Computation and Language · Computer Science 2025-03-25 Sharan Maiya , Yinhong Liu , Ramit Debnath , Anna Korhonen

Probing has become a go-to methodology for interpreting and analyzing deep neural models in natural language processing. However, there is still a lack of understanding of the limitations and weaknesses of various types of probes. In this…

Computation and Language · Computer Science 2022-11-15 Afra Amini , Tiago Pimentel , Clara Meister , Ryan Cotterell

World models learn to simulate environment dynamics from experience, enabling sample-efficient reinforcement learning. But what do these models actually represent internally? We apply interpretability techniques--including linear and…

Machine Learning · Computer Science 2026-03-24 Xinyu Zhang

While large language models (LLMs) excel in mathematical and code reasoning, we observe they struggle with social reasoning tasks, exhibiting cognitive confusion, logical inconsistencies, and conflation between objective world states and…

Computation and Language · Computer Science 2025-10-14 Jialu Du , Guiyang Hou , Yihui Fu , Chen Wu , Wenqi Zhang , Yongliang Shen , Weiming Lu

Large language models (LLMs) trained purely on text ostensibly lack any direct perceptual experience, yet their internal representations are implicitly shaped by multimodal regularities encoded in language. We test the hypothesis that…

Computation and Language · Computer Science 2025-10-06 Sophie L. Wang , Phillip Isola , Brian Cheung

Evaluation of biases in language models is often limited to synthetically generated datasets. This dependence traces back to the need for a prompt-style dataset to trigger specific behaviors of language models. In this paper, we address…

Computation and Language · Computer Science 2022-05-16 Sarah Alnegheimish , Alicia Guo , Yi Sun

Although neural models have achieved impressive results on several NLP benchmarks, little is understood about the mechanisms they use to perform language tasks. Thus, much recent attention has been devoted to analyzing the sentence…

Computation and Language · Computer Science 2021-03-09 Abhilasha Ravichander , Yonatan Belinkov , Eduard Hovy

Eliciting Latent Knowledge (ELK) aims to find patterns in a capable neural network's activations that robustly track the true state of the world, especially in hard-to-verify cases where the model's output is untrusted. To further ELK…

Machine Learning · Computer Science 2024-08-12 Alex Mallen , Madeline Brumley , Julia Kharchenko , Nora Belrose

Large Language Models (LLMs) are prone to generating fluent but incorrect content, known as confabulation, which poses increasing risks in multi-turn or agentic applications where outputs may be reused as context. In this work, we…

Computation and Language · Computer Science 2026-03-18 Tianyi Zhou , Johanne Medina , Sanjay Chawla

Progress in pre-trained language models has led to a surge of impressive results on downstream tasks for natural language understanding. Recent work on probing pre-trained language models uncovered a wide range of linguistic properties…

Computation and Language · Computer Science 2022-03-22 Zeming Chen , Qiyue Gao

Probing is widely used to study which features can be decoded from language model representations. However, the common decoding probe approach has two limitations that we aim to solve with our new encoding probe approach: contributions of…

Computation and Language · Computer Science 2026-05-04 Gaofei Shen , Martijn Bentum , Tom Lentz , Afra Alishahi , Grzegorz Chrupała

Probing is a popular method to discern what linguistic information is contained in the representations of pre-trained language models. However, the mechanism of selecting the probe model has recently been subject to intense debate, as it is…

Computation and Language · Computer Science 2022-07-06 Jiaoda Li , Ryan Cotterell , Mrinmaya Sachan

Do large language models (LLMs) exhibit sociodemographic biases, even when they decline to respond? To bypass their refusal to "speak," we study this research question by probing contextualized embeddings and exploring whether this bias is…

Computation and Language · Computer Science 2023-12-01 Raphael Tang , Xinyu Zhang , Jimmy Lin , Ferhan Ture

We investigate how language models handle the proviso problem, an unresolved issue in pragmatics where presuppositions in conditional sentences diverge between theoretical and human interpretations. We reformulate this phenomenon as a…

Computation and Language · Computer Science 2026-03-10 Tara Azin , Daniel Dumitrescu , Diana Inkpen , Raj Singh

Recent work has demonstrated that the latent spaces of large language models (LLMs) contain directions predictive of the truth of sentences. Multiple methods recover such directions and build probes that are described as uncovering a…

Computation and Language · Computer Science 2025-07-14 Stefan F. Schouten , Peter Bloem , Ilia Markov , Piek Vossen

Bias research in NLP seeks to analyse models for social biases, thus helping NLP practitioners uncover, measure, and mitigate social harms. We analyse the body of work that uses prompts and templates to assess bias in language models. We…

Computation and Language · Computer Science 2023-05-23 Seraphina Goldfarb-Tarrant , Eddie Ungless , Esma Balkir , Su Lin Blodgett

Large language models (LLMs) contain substantial factual knowledge which is commonly elicited by multiple-choice question-answering prompts. Internally, such models process the prompt through multiple transformer layers, building varying…

Computation and Language · Computer Science 2025-01-31 Didier Chételat , Joseph Cotnareanu , Rylee Thompson , Yingxue Zhang , Mark Coates

Prompt-based probing has been widely used in evaluating the abilities of pretrained language models (PLMs). Unfortunately, recent studies have discovered such an evaluation may be inaccurate, inconsistent and unreliable. Furthermore, the…

Computation and Language · Computer Science 2022-03-24 Boxi Cao , Hongyu Lin , Xianpei Han , Fangchao Liu , Le Sun

Accounts of human language processing have long appealed to implicit ``situation models'' that enrich comprehension with relevant but unstated world knowledge. Here, we apply causal intervention techniques to recent transformer models to…

Computation and Language · Computer Science 2023-06-08 Takateru Yamakoshi , James L. McClelland , Adele E. Goldberg , Robert D. Hawkins