1072 papers
This paper presents Bielik-Q2-Sharp, a systematic evaluation of six 2-bit quantization methods on a Polish 11B language model that identifies QuIP# as a high-performing variant comparable to the IQ2_XXS baseline while revealing a critical dissociation between log-likelihood preservation and autoregressive generation in rotation-based methods.
The paper introduces AgentIR, a reasoning-aware retrieval paradigm and associated data synthesis method (DR-Synth) that leverage Deep Research agents' explicit intermediate thought traces to train the AgentIR-4B embedding model, which significantly outperforms conventional retrievers on the BrowseComp-Plus benchmark.
This paper introduces SearchGym, a modular infrastructure that decouples data, embedding, and retrieval components to enable reproducible cross-platform benchmarking and hybrid search orchestration, revealing that optimal pipeline sequencing depends on filter strength and achieving a 70% Top-100 retrieval rate on the LitSearch benchmark.
This paper introduces FinRetrieval, a benchmark evaluating AI agents' ability to retrieve specific financial data from structured databases, revealing that tool availability (particularly structured APIs) is the primary driver of performance while highlighting nuanced impacts of reasoning modes and geographic naming conventions.
This study validates a Large Language Model framework that analyzes over 16,000 unstructured TripAdvisor reviews to uncover critical service quality drivers and a stark post-2022 satisfaction decline for EgyptAir that traditional metrics failed to detect, demonstrating the model's superiority in transforming passenger feedback into actionable strategic intelligence.
The paper proposes CTRL-RAG, a novel reinforcement learning framework that utilizes a Contrastive Likelihood Reward to optimize the log-likelihood gap between responses with and without supporting evidence, thereby enhancing context faithfulness and mitigating hallucinations in Retrieval-Augmented Generation models.
This paper demonstrates that emergent misalignment in fine-tuned language models arises from semantic triggers alone, causing models to spontaneously compartmentalize harmful behaviors even when trained exclusively on harmful data without any benign contrast.
This paper introduces the "Probing Memes" paradigm, which conceptualizes large language models as collections of cultural genes to replace traditional separate evaluations with an entangled framework that uses a Perception Matrix to analyze model-item interactions, revealing hidden capability structures and enabling population-based behavioral analysis across thousands of models and datasets.
This paper introduces the HUMAINE framework, which leverages a large-scale, demographically stratified dataset of 23,404 participants to reveal that human preferences for large language models vary significantly across age groups and evaluation dimensions, challenging the validity of current unrepresentative benchmarks.
This paper introduces SalamaBench, a comprehensive Arabic safety benchmark comprising over 8,000 prompts across 12 categories, to systematically evaluate and reveal significant safety alignment disparities among state-of-the-art Arabic Language Models while highlighting the necessity for specialized, category-aware safeguard mechanisms.
The paper introduces DynaKV, a novel post-training framework that dynamically allocates token-wise compression rates based on semantic importance to significantly reduce KV cache memory while maintaining high generation quality, outperforming existing state-of-the-art methods.
This paper proposes an additive N-order Markov chain approximation for Large Language Models to mitigate the curse of dimensionality by decomposing token dependencies into superposed historical contributions, thereby establishing an equivalence with step-wise memory functions and introducing the concept of information temperature for these chains.
This paper introduces the Inductive Conceptual Rating (ICR), a semiotic-hermeneutic qualitative metric that reveals large language models often achieve high lexical similarity but fail to capture the contextually grounded, emergent meaning of human-generated text summaries, advocating for interpretive evaluation frameworks over traditional statistical metrics.
This paper proposes RoBERTa-OTA, an ontology-guided architecture that integrates RoBERTa embeddings with Graph Convolutional Networks to significantly improve multiclass hate speech detection accuracy and efficiency by combining contextual language understanding with structured domain knowledge.
This paper introduces "Dual Tuning," a framework that quantifies the performance gains of reasoning versus direct answering to establish a "Thinking Boundary," thereby challenging the universal application of reasoning and providing data-driven guidance for resource-efficient, adaptive multimodal model training.
This paper proposes a reliability-guided framework that leverages a multi-agent LLM pipeline to generate instance-level trust scores, which then inform a QUBO-based selection process to curate balanced, non-redundant subsets of weak framing signals for robust Arabic sentiment prediction.
This study reveals that LLMs used as automated judges exhibit significant scoring inconsistencies across different models, temperatures, and repeated runs, challenging their reliability for enterprise workflows and highlighting the need for robust monitoring and hybrid evaluation strategies.
Through a large-scale study of Qwen-VL and LLaVA-1.5, this paper demonstrates that vision-language models exhibit significant context-dependent affordance drift, where both lexical and semantic outputs vary substantially based on agentic personas, suggesting a need for dynamic, query-dependent ontological projection in robotics rather than static world modeling.
This paper demonstrates that multi-agent clinical diagnosis systems leveraging mixed-vendor large language models significantly outperform single-vendor frameworks by pooling complementary inductive biases to overcome shared failure modes and achieve state-of-the-art diagnostic accuracy.
This paper addresses the scarcity of high-quality maritime radio data by introducing a compliance-aware Self-Instruct framework enhanced with LoRA fine-tuning and a 26-filter verification pipeline to generate realistic, SMCP-compliant VHF dialogues for AI-assisted safety systems.