English
Related papers

Related papers: Eliciting Latent Knowledge from Quirky Language Mo…

200 papers

As language models (LMs) deliver increasing performance on a range of NLP tasks, probing classifiers have become an indispensable technique in the effort to better understand their inner workings. A typical setup involves (1) defining an…

Computation and Language · Computer Science 2024-08-01 Charles Jin , Martin Rinard

Can large language models (LLMs) express their uncertainty in situations where they lack sufficient parametric knowledge to generate reasonable responses? This work aims to systematically investigate LLMs' behaviors in such situations,…

Computation and Language · Computer Science 2024-02-19 Genglin Liu , Xingyao Wang , Lifan Yuan , Yangyi Chen , Hao Peng

Previous literatures show that pre-trained masked language models (MLMs) such as BERT can achieve competitive factual knowledge extraction performance on some datasets, indicating that MLMs can potentially be a reliable knowledge source. In…

Computation and Language · Computer Science 2021-06-18 Boxi Cao , Hongyu Lin , Xianpei Han , Le Sun , Lingyong Yan , Meng Liao , Tong Xue , Jin Xu

One of the major aspects contributing to the striking performance of large language models (LLMs) is the vast amount of factual knowledge accumulated during pre-training. Yet, many LLMs suffer from self-inconsistency, which raises doubts…

Computation and Language · Computer Science 2024-10-07 Anastasiia Sedova , Robert Litschko , Diego Frassinelli , Benjamin Roth , Barbara Plank

Large language models (LLMs) rarely admit uncertainty, often producing fluent but misleading answers, rather than abstaining (i.e., refusing to answer). This weakness is even evident in temporal question answering, where models frequently…

Computation and Language · Computer Science 2026-03-05 Xinyu Zhou , Chang Jin , Carsten Eickhoff , Zhijiang Guo , Seyed Ali Bahrainian

Temporary syntactic ambiguities arise when the beginning of a sentence is compatible with multiple syntactic analyses. We inspect to which extent neural language models (LMs) exhibit uncertainty over such analyses when processing…

Computation and Language · Computer Science 2021-09-17 Laura Aina , Tal Linzen

Knowledge probing quantifies how much relational knowledge a language model (LM) has acquired during pre-training. Existing knowledge probes evaluate model capabilities through metrics like prediction accuracy and precision. Such…

Computation and Language · Computer Science 2026-01-28 Christopher Kissling , Elena Merdjanovska , Alan Akbik

Large language models (LLMs) sometimes demonstrate poor performance on knowledge-intensive tasks, commonsense reasoning is one of them. Researchers typically address these issues by retrieving related knowledge from knowledge graphs or…

Computation and Language · Computer Science 2024-10-15 Jiachun Li , Pengfei Cao , Chenhao Wang , Zhuoran Jin , Yubo Chen , Kang Liu , Xiaojian Jiang , Jiexin Xu , Jun Zhao

Large language models (LLMs) can generate long-form and coherent text, yet they often hallucinate facts, which undermines their reliability. To mitigate this issue, inference-time methods steer LLM representations toward the "truthful…

Computation and Language · Computer Science 2024-06-10 Farima Fatahi Bayat , Xin Liu , H. V. Jagadish , Lu Wang

Automated fact-checking has been a challenging task for the research community. Prior work has explored various strategies, such as end-to-end training, retrieval-augmented generation, and prompt engineering, to build robust fact-checking…

Computation and Language · Computer Science 2026-02-23 Gaurav Kumar , Ayush Garg , Debajyoti Mazumder , Aditya Kishore , Babu kumar , Jasabanta Patro

Language models are susceptible to bias, sycophancy, backdoors, and other tendencies that lead to unfaithful responses to the input context. Interpreting internal states of language models could help monitor and correct unfaithful behavior.…

Computation and Language · Computer Science 2024-12-10 Jiahai Feng , Stuart Russell , Jacob Steinhardt

Intrinsic self-correct was a method that instructed large language models (LLMs) to verify and correct their responses without external feedback. Unfortunately, the study concluded that the LLMs could not self-correct reasoning yet. We find…

Computation and Language · Computer Science 2024-10-04 Zhenyu Wu , Qingkai Zeng , Zhihan Zhang , Zhaoxuan Tan , Chao Shen , Meng Jiang

Large Language Models (LLMs) perform well in general QA but often struggle in domain-specific scenarios. Retrieval-Augmented Generation (RAG) introduces external knowledge but suffers from hallucinations and latency due to noisy retrievals.…

Computation and Language · Computer Science 2025-09-19 Bolei He , Xinran He , Run Shao , Shanfu Shu , Xianwei Xue , Mingquan Cheng , Haifeng Li , Zhenhua Ling

Knowledge distillation from large language models (LLMs) assumes that the teacher's output distribution is a high-quality training signal. On reasoning tasks, this assumption is frequently violated. A model's intermediate representations…

Computation and Language · Computer Science 2026-03-16 Ryan Brown , Chris Russell

Linear probes can detect when language models produce outputs they "know" are wrong, a capability relevant to both deception and reward hacking. However, single-layer probes are fragile: the best layer varies across models and tasks, and…

Machine Learning · Computer Science 2026-04-16 Erik Nordby , Tasha Pais , Aviel Parrack

The reliability of large language models (LLMs) is greatly compromised by their tendency to hallucinate, underscoring the need for precise identification of knowledge gaps within LLMs. Various methods for probing such gaps exist, ranging…

Computation and Language · Computer Science 2025-06-02 Raoyuan Zhao , Abdullatif Köksal , Ali Modarressi , Michael A. Hedderich , Hinrich Schütze

Deep directed generative models have attracted much attention recently due to their expressive representation power and the ability of ancestral sampling. One major difficulty of learning directed models with many latent variables is the…

Machine Learning · Computer Science 2015-06-16 Siqi Nie , Qiang Ji

LLMs often produce fluent but incorrect answers, yet detecting such hallucinations typically requires multiple sampling passes or post-hoc verification, adding significant latency and cost. We hypothesize that intermediate layers encode…

Computation and Language · Computer Science 2026-01-30 Rohan Bhatnagar , Youran Sun , Chi Andrew Zhang , Yixin Wen , Haizhao Yang

Knowledge editing methods (KEs) can update language models' obsolete or inaccurate knowledge learned from pre-training. However, KEs can be used for malicious applications, e.g., inserting misinformation and toxic content. Knowing whether a…

Computation and Language · Computer Science 2025-02-11 Paul Youssef , Zhixue Zhao , Christin Seifert , Jörg Schlötterer

Despite their impressive capabilities, Large Language Models (LLMs) exhibit unwanted uncertainty, a phenomenon where a model changes a previously correct answer into an incorrect one when re-prompted. This behavior undermines trust and…

Computation and Language · Computer Science 2025-10-28 Tiasa Singha Roy , Ayush Rajesh Jhaveri , Ilias Triantafyllopoulos