mailitics

Category: cs.CL

  • Deep networks learn to parse uniform-depth context-free languages from local statistics

    Deep networks learn to parse uniform-depth context-free languages from local statistics arXiv:2602.06065v1 Announce Type: new Abstract: Understanding how the structure of language can be learned from sentences alone is a central question in both cognitive science and machine learning. Studies of the internal representations of Large Language Models (LLMs) support their ability to parse text…

  • Towards Latent Diffusion Suitable For Text

    Towards Latent Diffusion Suitable For Text arXiv:2601.16220v1 Announce Type: cross Abstract: Language diffusion models aim to improve sampling speed and coherence over autoregressive LLMs. We introduce Neural Flow Diffusion Models for language generation, an extension of NFDM that enables the straightforward application of continuous diffusion models to discrete state spaces. NFDM learns a multivariate forward…

  • A path to natural language through tokenisation and transformers

    A path to natural language through tokenisation and transformers arXiv:2601.03368v1 Announce Type: cross Abstract: Natural languages exhibit striking regularities in their statistical structure, including notably the emergence of Zipf’s and Heaps’ laws. Despite this, it remains broadly unclear how these properties relate to the modern tokenisation schemes used in contemporary transformer models. In this note,…

  • Detecting and Mitigating Treatment Leakage in Text-Based Causal Inference: Distillation and Sensitivity Analysis

    Detecting and Mitigating Treatment Leakage in Text-Based Causal Inference: Distillation and Sensitivity Analysis arXiv:2601.02400v1 Announce Type: cross Abstract: Text-based causal inference increasingly employs textual data as proxies for unobserved confounders, yet this approach introduces a previously undertheorized source of bias: treatment leakage. Treatment leakage occurs when text intended to capture confounding information also contains signals…

  • Bayesian Evaluation of Large Language Model Behavior

    Bayesian Evaluation of Large Language Model Behavior arXiv:2511.10661v1 Announce Type: cross Abstract: It is increasingly important to evaluate how text generation systems based on large language models (LLMs) behave, such as their tendency to produce harmful output or their sensitivity to adversarial inputs. Such evaluations often rely on a curated benchmark set of input prompts…

  • The Coverage Principle: How Pre-training Enables Post-Training

    The Coverage Principle: How Pre-training Enables Post-Training arXiv:2510.15020v1 Announce Type: new Abstract: Language models demonstrate remarkable abilities when pre-trained on large text corpora and fine-tuned for specific tasks, but how and why pre-training shapes the success of the final model remains poorly understood. Notably, although pre-training success is often quantified by cross entropy loss, cross-entropy…

  • Domain-Shift-Aware Conformal Prediction for Large Language Models

    Domain-Shift-Aware Conformal Prediction for Large Language Models arXiv:2510.05566v1 Announce Type: new Abstract: Large language models have achieved impressive performance across diverse tasks. However, their tendency to produce overconfident and factually incorrect outputs, known as hallucinations, poses risks in real world applications. Conformal prediction provides finite-sample, distribution-free coverage guarantees, but standard conformal prediction breaks down under…

  • BED-LLM: Intelligent Information Gathering with LLMs and Bayesian Experimental Design

    BED-LLM: Intelligent Information Gathering with LLMs and Bayesian Experimental Design arXiv:2508.21184v1 Announce Type: cross Abstract: We propose a general-purpose approach for improving the ability of Large Language Models (LLMs) to intelligently and adaptively gather information from a user or other external source using the framework of sequential Bayesian experimental design (BED). This enables LLMs to…

  • On the Fundamental Impossibility of Hallucination Control in Large Language Models

    On the Fundamental Impossibility of Hallucination Control in Large Language Models arXiv:2506.06382v1 Announce Type: new Abstract: This paper explains textbf{why it is impossible to create large language models that do not hallucinate and what are the trade-offs we should be looking for}. It presents a formal textbf{impossibility theorem} demonstrating that no inference mechanism can simultaneously…

  • Boosting In-Context Learning in LLMs Through the Lens of Classical Supervised Learning

    Boosting In-Context Learning in LLMs Through the Lens of Classical Supervised Learning arXiv:2505.23783v1 Announce Type: new Abstract: In-Context Learning (ICL) allows Large Language Models (LLMs) to adapt to new tasks with just a few examples, but their predictions often suffer from systematic biases, leading to unstable performances in classification. While calibration techniques are proposed to…

  • Cer-Eval: Certifiable and Cost-Efficient Evaluation Framework for LLMs

    Cer-Eval: Certifiable and Cost-Efficient Evaluation Framework for LLMs arXiv:2505.03814v1 Announce Type: new Abstract: As foundation models continue to scale, the size of trained models grows exponentially, presenting significant challenges for their evaluation. Current evaluation practices involve curating increasingly large datasets to assess the performance of large language models (LLMs). However, there is a lack of…

  • (Im)possibility of Automated Hallucination Detection in Large Language Models

    (Im)possibility of Automated Hallucination Detection in Large Language Models arXiv:2504.17004v1 Announce Type: cross Abstract: Is automated hallucination detection possible? In this work, we introduce a theoretical framework to analyze the feasibility of automatically detecting hallucinations produced by large language models (LLMs). Inspired by the classical Gold-Angluin framework for language identification and its recent adaptation to…

  • How Private is Your Attention? Bridging Privacy with In-Context Learning

    How Private is Your Attention? Bridging Privacy with In-Context Learning arXiv:2504.16000v1 Announce Type: new Abstract: In-context learning (ICL)-the ability of transformer-based models to perform new tasks from examples provided at inference time-has emerged as a hallmark of modern language models. While recent works have investigated the mechanisms underlying ICL, its feasibility under formal privacy constraints…

  • StealthRank: LLM Ranking Manipulation via Stealthy Prompt Optimization

    StealthRank: LLM Ranking Manipulation via Stealthy Prompt Optimization arXiv:2504.05804v1 Announce Type: cross Abstract: The integration of large language models (LLMs) into information retrieval systems introduces new attack surfaces, particularly for adversarial ranking manipulations. We present StealthRank, a novel adversarial ranking attack that manipulates LLM-driven product recommendation systems while maintaining textual fluency and stealth. Unlike existing…

  • Towards Interpretable Soft Prompts

    Towards Interpretable Soft Prompts arXiv:2504.02144v1 Announce Type: cross Abstract: Soft prompts have been popularized as a cheap and easy way to improve task-specific LLM performance beyond few-shot prompts. Despite their origin as an automated prompting method, however, soft prompts and other trainable prompts remain a black-box method with no immediately interpretable connections to prompting. We…

  • Fundamental Safety-Capability Trade-offs in Fine-tuning Large Language Models

    Fundamental Safety-Capability Trade-offs in Fine-tuning Large Language Models arXiv:2503.20807v1 Announce Type: new Abstract: Fine-tuning Large Language Models (LLMs) on some task-specific datasets has been a primary use of LLMs. However, it has been empirically observed that this approach to enhancing capability inevitably compromises safety, a phenomenon also known as the safety-capability trade-off in LLM fine-tuning.…

  • An Overview of Large Language Models for Statisticians

    An Overview of Large Language Models for Statisticians arXiv:2502.17814v1 Announce Type: new Abstract: Large Language Models (LLMs) have emerged as transformative tools in artificial intelligence (AI), exhibiting remarkable capabilities across diverse tasks such as text generation, reasoning, and decision-making. While their success has primarily been driven by advances in computational power and deep learning architectures,…

  • Circuit Complexity Bounds for Visual Autoregressive Model

    Circuit Complexity Bounds for Visual Autoregressive Model arXiv:2501.04299v1 Announce Type: new Abstract: Understanding the expressive ability of a specific model is essential for grasping its capacity limitations. Recently, several studies have established circuit complexity bounds for Transformer architecture. Besides, the Visual AutoRegressive (VAR) model has risen to be a prominent method in the field of…

  • Who Wrote This? Zero-Shot Statistical Tests for LLM-Generated Text Detection using Finite Sample Concentration Inequalities

    Who Wrote This? Zero-Shot Statistical Tests for LLM-Generated Text Detection using Finite Sample Concentration Inequalities arXiv:2501.02406v1 Announce Type: new Abstract: Verifying the provenance of content is crucial to the function of many organizations, e.g., educational institutions, social media platforms, firms, etc. This problem is becoming increasingly difficult as text generated by Large Language Models (LLMs)…

  • Many of Your DPOs are Secretly One: Attempting Unification Through Mutual Information

    Many of Your DPOs are Secretly One: Attempting Unification Through Mutual Information arXiv:2501.01544v1 Announce Type: cross Abstract: Post-alignment of large language models (LLMs) is critical in improving their utility, safety, and alignment with human intentions. Direct preference optimisation (DPO) has become one of the most widely used algorithms for achieving this alignment, given its ability…

  • How to Choose a Threshold for an Evaluation Metric for Large Language Models

    How to Choose a Threshold for an Evaluation Metric for Large Language Models arXiv:2412.12148v1 Announce Type: new Abstract: To ensure and monitor large language models (LLMs) reliably, various evaluation metrics have been proposed in the literature. However, there is little research on prescribing a methodology to identify a robust threshold on these metrics even though…

  • Training-Free Bayesianization for Low-Rank Adapters of Large Language Models

    Training-Free Bayesianization for Low-Rank Adapters of Large Language Models arXiv:2412.05723v1 Announce Type: new Abstract: Estimating the uncertainty of responses of Large Language Models~(LLMs) remains a critical challenge. While recent Bayesian methods have demonstrated effectiveness in quantifying uncertainty through low-rank weight updates, they typically require complex fine-tuning or post-training procedures. In this paper, we propose Training-Free…

  • Composition of Experts: A Modular Compound AI System Leveraging Large Language Models

    Composition of Experts: A Modular Compound AI System Leveraging Large Language Models arXiv:2412.01868v1 Announce Type: cross Abstract: Large Language Models (LLMs) have achieved remarkable advancements, but their monolithic nature presents challenges in terms of scalability, cost, and customization. This paper introduces the Composition of Experts (CoE), a modular compound AI system leveraging multiple expert LLMs.…