English
Related papers

Related papers: Correcting Suppressed Log-Probabilities in Languag…

200 papers

While the numerous parameters in Large Language Models (LLMs) contribute to their superior performance, this massive scale makes them inefficient and memory-hungry. Thus, they are hard to deploy on commodity hardware, such as one single…

Computation and Language · Computer Science 2023-10-11 Zhaozhuo Xu , Zirui Liu , Beidi Chen , Yuxin Tang , Jue Wang , Kaixiong Zhou , Xia Hu , Anshumali Shrivastava

We uncover a latent capacity for introspection in a Qwen 32B model, demonstrating that the model can detect when concepts have been injected into its earlier context and identify which concept was injected. While the model denies injection…

Artificial Intelligence · Computer Science 2026-02-27 Theia Pearson-Vogel , Martin Vanek , Raymond Douglas , Jan Kulveit

In this paper, we introduce a novel technique for content safety and prompt injection classification for Large Language Models. Our technique, Layer Enhanced Classification (LEC), trains a Penalized Logistic Regression (PLR) classifier on…

Computation and Language · Computer Science 2024-12-19 Mason Sawtell , Tula Masterman , Sandi Besen , Jim Brown

Embedding layers in transformer-based NLP models typically account for the largest share of model parameters, scaling with vocabulary size but not yielding performance gains proportional to scale. We propose an alternative approach in which…

Computation and Language · Computer Science 2025-05-06 Henry Ndubuaku , Mouad Talhi

Recent work (e.g. LAMA (Petroni et al., 2019)) has found that the quality of the factual information extracted from Large Language Models (LLMs) depends on the prompts used to query them. This inconsistency is problematic because different…

Computation and Language · Computer Science 2022-04-20 Benjamin Newman , Prafulla Kumar Choubey , Nazneen Rajani

Pretrained language models (LMs) are prone to arithmetic errors. Existing work showed limited success in probing numeric values from models' representations, indicating that these errors can be attributed to the inherent unreliability of…

Computation and Language · Computer Science 2025-10-27 Marek Kadlčík , Michal Štefánik , Timothee Mickus , Michal Spiegel , Josef Kuchař

Large language models are increasingly deployed in high-stakes tasks, where confident yet incorrect inferences may cause severe real-world harm, bringing the previously overlooked issue of confidence faithfulness back to the forefront. A…

Machine Learning · Computer Science 2026-04-10 Haokai Ma , Lee Yan Zhen , Gang Yang , Yunshan Ma , Ee-Chien Chang , Tat-Seng Chua

Large vision-language models (VLMs) often rely on familiar semantic priors, but existing evaluations do not cleanly separate perception failures from rule-mapping failures. We study this behavior as semantic fixation: preserving a default…

Computer Vision and Pattern Recognition · Computer Science 2026-04-15 Md Tanvirul Alam

Transformer attention computes a single softmax-weighted average over values -- a one-pass estimate that cannot correct its own errors. We introduce \emph{gradient-boosted attention}, which applies the principle of gradient boosting…

Machine Learning · Computer Science 2026-05-05 Saleh Sargolzaei

Recent advancements in language models (LMs) have marked a shift toward the growing importance of post-training. Yet, post-training approaches such as supervised fine-tuning (SFT) do not guarantee the effective use of knowledge acquired…

Computation and Language · Computer Science 2025-10-30 Chunyuan Deng , Ruidi Chang , Hanjie Chen

LLMs reliably correct false claims when presented in isolation, yet when the same claims are embedded in task-oriented requests, they often comply rather than correct. We term this failure mode \emph{correction suppression} and construct a…

Machine Learning · Computer Science 2026-05-11 Zixuan Chen , Hao Lin , Zizhe Chen , Yizhou Tian , Garry Yang , Depeng Wang , Ya Guo , Huijia Zhu , James Cheng

As large language models (LLMs) rapidly displace traditional expertise, their capacity to correct misinformation has become a core concern. We investigate the idea that prompt framing systematically modulates misinformation correction -…

Human-Computer Interaction · Computer Science 2025-12-01 Sekoul Krastev , Hilary Sweatman , Anni Sternisko , Steve Rathje

The ability of machine learning models to store input information in hidden layer vector embeddings, analogous to the concept of `memory', is widely employed but not well characterized. We find that language model embeddings typically…

Computation and Language · Computer Science 2026-05-20 Benjamin L. Badger

Complex problems, whether in math, logic, or planning, are solved by humans through a sequence of steps where the result of one step informs the next. In this work, we adopt the perspective that the reasoning power of Transformers is…

Machine Learning · Computer Science 2026-02-23 Mohan Tang , Sidi Lu

Although state-of-the-art LLMs can solve math problems, we find that they make errors on numerical comparisons with mixed notation: "Which is larger, $5.7 \times 10^2$ or $580$?" This raises a fundamental question: Do LLMs even know how big…

Computation and Language · Computer Science 2026-02-18 Fengting Yuchi , Li Du , Jason Eisner

In-context learning (ICL) allows Transformers to adapt to novel tasks without weight updates, yet the underlying algorithms remain poorly understood. We adopt a statistical decision-theoretic perspective by investigating simple binary…

Machine Learning · Computer Science 2026-03-13 Faris Chaudhry , Siddhant Gadkari

Large language models (LLMs) are increasingly deployed with task-specific adapters catering to multiple downstream applications. In such a scenario, the additional compute associated with these apparently insignificant number of adapter…

Computation and Language · Computer Science 2025-10-31 Dhananjaya Gowda , Seoha Song , Harshith Goka , Junhyun Lee

Pre-trained transformer models shine in many natural language processing tasks and therefore are expected to bear the representation of the input sentence or text meaning. These sentence-level embeddings are also important in…

Computation and Language · Computer Science 2025-02-21 Lukas Stankevičius , Mantas Lukoševičius

Chain-of-Thought (CoT) prompting has significantly enhanced the reasoning abilities of large language models. However, recent studies have shown that models can still perform complex reasoning tasks even when the CoT is replaced with…

Computation and Language · Computer Science 2024-12-09 Aryasomayajula Ram Bharadwaj

This research aims to unravel how large language models (LLMs) iteratively refine token predictions through internal processing. We utilized a logit lens technique to analyze the model's token predictions derived from intermediate…

Computation and Language · Computer Science 2025-06-10 Jaturong Kongmanee