The paper introduces an automated framework demonstrating that LLM system instructions are vulnerable to encoding attacks, where structured output requests can bypass safety refusals and leak sensitiv…

This paper evaluates the causal reasoning abilities of large language models and finds that they rely heavily on lexical pattern matching rather than structural reasoning.

The paper introduces ACROS, a method that induces an explicit sense representation pathway into a frozen pretrained decoder LM, enabling sense-based tasks like disambiguation and cross-lingual alignme…

The paper introduces SONAR, a prompt sanitization framework that uses natural language inference metrics to identify and remove malicious instructions injected into LLM prompts, achieving near-zero at…

The paper introduces FRANZ, a communicative audit framework, to evaluate how LLMs frame responses to subjective questions, finding that LLMs exhibit statistically significant and coupled differences i…

GrepSeek introduces a novel direct corpus interaction (DCI) search agent that trains an LLM to find and compose evidence from large text corpora by issuing executable shell commands, achieving state-o…

This paper investigates how the final prompt in conversational AI-search evaluations differs from the conversation history, using two corpora of commercial and PRISM conversations.

This paper compares the preservation of linguistic content in different text encodings (tokens, bytes, pixels) using a shared bottleneck, revealing their distinct strengths in surface form preservatio…

SentGuard introduces a novel sentence-level streaming guardrail that operates in parallel with LLM generation, achieving high detection rates of unsafe content early in the response while maintaining…

This paper compares three methods for answering questions over legal documents: full-corpus injection, embedding retrieval, and LLM navigation over a compact structured index. The results show that th…

This paper proposes a multi-turn retrieval-augmented generation pipeline for conversational systems across four domains.

This paper reports controlled experiments on prompt-design decisions for instruction-following and context length in AI models, finding significant degradation in performance beyond certain thresholds…

The paper compares verbalized feature attributions and self-generated rationales for explaining model behavior, finding that the format and granularity of the explanation significantly affect its abil…

The paper introduces Semantic Triplet Restoration (STR), a novel protocol that converts complex table structures into atomic semantic triplets, improving table question answering by providing explicit…

This paper explores the effect of conversational timing properties on automatic speech recognition (ASR) systems by controlling and optimizing pause and overlap timing distributions.

The paper proposes SubFit, a novel compression technique that achieves superior LLM compression by replacing non-contiguous, submodule-level components (Attention and FeedForward) with lightweight res…

HypothesisMed introduces an inference-time pipeline for biomedical question answering that improves model reliability and structured output generation by fusing multiple model outputs and diagnosing t…

The paper introduces OCC-RAG, a family of compact, task-specialized Small Language Models (SLMs) designed to achieve highly faithful, multi-hop question answering grounded strictly in provided context…

The paper introduces MIDI, a novel multilingual dataset that embeds idioms in realistic sentence and conversational contexts across diverse resource levels, revealing that idiom comprehension is signi…

This paper presents P4IR, a two-stage framework for automated code compliance using supervised fine-tuning and Group Relative Policy Optimization, achieving significant improvements over baselines and…