Related papers: Beyond the Covariance Trap: Unlocking Generalizati…

Large Language Models (LLMs) have recently demonstrated strong capabilities in code-related tasks, but their robustness in code reasoning under perturbations remains underexplored. We introduce CodeCrash, a stress-testing framework with…

Accurate and unambiguous guidelines are critical for large language model (LLM) based graders, yet manually crafting these prompts is often sub-optimal as LLMs can misinterpret expert guidelines or lack necessary domain specificity.…

Large Reasoning Models (LRMs) generate structured chains of thought (CoTs) before producing final answers, making them especially vulnerable to knowledge leakage through intermediate reasoning steps. Yet, the memorization of sensitive…

Reasoning over structured graphs remains a fundamental challenge for Large Language Models (LLMs), particularly when scaling to large graphs. Existing approaches typically follow the retrieval-augmented generation (RAG) paradigm: first…

As real-world knowledge evolves, the information embedded within large language models (LLMs) can become outdated, inadequate, or erroneous. Model editing has emerged as a prominent approach for updating LLMs' knowledge with minimal…

Language models typically tokenize raw text into sequences of subword identifiers from a predefined vocabulary, a process inherently sensitive to typographical errors, length variations, and largely oblivious to the internal structure of…

Large language models (LLMs) acquire vast knowledge from large text corpora, but this information can become outdated or inaccurate. Since retraining is computationally expensive, knowledge editing offers an efficient alternative --…

Neural collapse describes the geometry of activation in the final layer of a deep neural network when it is trained beyond performance plateaus. Open questions include whether neural collapse leads to better generalization and, if so, why…

Model editing is an emerging field that focuses on updating the knowledge embedded within large language models (LLMs) without extensive retraining. However, current model editing methods significantly compromise the general abilities of…

Large Language Models (LLMs) often suffer from catastrophic forgetting and collapse during sequential knowledge editing. This vulnerability stems from the prevailing dense editing paradigm, which treats models as black boxes and relies on…

Knowledge editing aims to update the embedded knowledge within Large Language Models (LLMs). However, existing approaches, whether through parameter modification or external memory integration, often suffer from inconsistent evaluation…

Large language models increasingly rely on synthetic data due to human-written content scarcity, yet recursive training on model-generated outputs leads to model collapse, a degenerative process threatening factual reliability. We define…

Large language models (LLMs) have demonstrated impressive capabilities and are receiving increasing attention to enhance their reasoning through scaling test--time compute. However, their application in open--ended, knowledge--intensive,…

Large Language Models (LLMs) are vulnerable to backdoor attacks that manipulate outputs via hidden triggers. Existing defense methods--designed for vision/text classification tasks--fail for text generation. We propose Internal Consistency…

Multimodal Large Language Models (MLLMs) have demonstrated remarkable capabilities in interpreting images using natural language. However, without using large-scale datasets for retraining, these models are difficult to adapt to specialized…

The integration of Large Language Models (LLMs) and Graph Neural Networks (GNNs) promises to unify semantic understanding with structural reasoning, yet existing methods typically rely on static, unidirectional pipelines. These approaches…

A prevailing narrative in LLM post-training holds that supervised finetuning (SFT) memorizes while reinforcement learning (RL) generalizes. We revisit this claim for reasoning SFT with long chain-of-thought (CoT) supervision and find that…

While reinforcement learning with verifiable rewards (RLVR) is effective to improve the reasoning ability of large language models (LLMs), its reliance on human-annotated labels leads to the scaling up dilemma, especially for complex tasks.…

Recent vision-language models (VLMs) achieve remarkable reasoning through reinforcement learning (RL), which provides a feasible solution for realizing continuous self-evolving large vision-language models (LVLMs) in the era of experience.…

Model editing aims to efficiently alter the behavior of Large Language Models (LLMs) within a desired scope, while ensuring no adverse impact on other inputs. Recent years have witnessed various model editing methods been proposed. However,…