12074 papers
This paper introduces AdAEM, a novel self-extensible evaluation framework that automatically generates adaptive test questions by probing the internal value boundaries of diverse LLMs to overcome the limitations of static benchmarks and provide more informative, distinguishable insights into models' value differences and alignment dynamics.
The paper introduces ESGenius, the first comprehensive benchmark comprising a curated corpus of authoritative ESG documents and a rigorously validated question-answer dataset, which reveals that while large language models exhibit moderate zero-shot performance in sustainability domains, their accuracy significantly improves when grounded in retrieval-augmented generation (RAG) using the provided source materials.
The paper introduces KramaBench, a comprehensive benchmark featuring 104 real-world data-to-insight challenges across diverse domains, which reveals that current AI systems struggle to orchestrate end-to-end data pipelines over data lakes, achieving a maximum of only 55% accuracy despite strong performance in isolated tasks.
This paper introduces VisioMath, a benchmark of 1,800 K-12 mathematics problems featuring visually similar diagrams, to reveal that current Large Multimodal Models struggle with fine-grained comparative reasoning due to image-text misalignment and to demonstrate that alignment-oriented strategies can significantly improve performance.
This paper critiques existing evaluations of LLM moral competence for over-relying on simplified scenarios and proposes a novel five-dimensional framework that reveals models often outperform humans in structured tasks but significantly underperform when required to discern moral relevance from noisy information, suggesting current assessments substantially overestimate their true moral reasoning capabilities.
This paper introduces ContextBench, a benchmark for evaluating methods that generate fluent inputs to trigger specific latent features in language models, and demonstrates that enhanced Evolutionary Prompt Optimization variants achieve state-of-the-art performance in balancing elicitation strength with linguistic fluency.
The paper introduces Sysformer, a novel approach that safeguards frozen large language models by learning to adapt system prompts in the embedding space to significantly improve safety robustness against harmful inputs and jailbreaking attacks without requiring costly model fine-tuning.
The paper proposes Iterative Quantum Feature Maps (IQFMs), a hybrid quantum-classical framework that constructs deep architectures by iteratively connecting shallow, noise-resilient quantum feature maps with classically computed weights to mitigate hardware limitations and achieve performance comparable to classical neural networks without optimizing variational quantum parameters.
SPARC introduces a novel framework that unifies concept representations across diverse AI architectures and modalities by enforcing global sparsity and cross-reconstruction loss, thereby creating a shared latent space that enables direct cross-model interpretability and applications like text-guided localization without manual alignment.
This paper proposes a unified instructional framework that integrates MOOCs, Smart Teaching, and AI into a coherent, teaching-driven pedagogy, formalizing them as a layered knowledge transformation model to maximize systemic educational potential through structured exposure, adaptive allocation, and efficiency amplification.
The paper introduces ExDD, a novel framework for industrial surface defect detection that overcomes data scarcity and uniform outlier assumptions by explicitly modeling dual feature distributions via parallel memory banks and generating context-aware synthetic defects using latent diffusion models.
This paper presents a multimodal large language model-based multi-agent system that significantly outperforms existing state-of-the-art methods in automatically extracting structured chemical information from diverse and complex literature graphics, thereby advancing AI-driven chemical research.
This paper introduces MAP, a training-free decoding method that mitigates hallucinations in Large Vision-Language Models by interpreting hidden states as a 2D semantic map and employing layer-wise criss-cross attention and global-local logit fusion to aggregate widely distributed factual information for improved factual consistency.
This paper introduces VLMQ, a post-training quantization framework tailored for vision-language models that leverages a gradient-driven importance factor to address visual over-representation and modality gaps, thereby achieving state-of-the-art performance across various model sizes and low-bit settings.
The paper proposes SGDFuse, a novel two-stage conditional diffusion model guided by Segment Anything Model (SAM) semantic masks, which achieves high-fidelity infrared and visible image fusion by leveraging explicit semantic priors to preserve key targets and minimize artifacts for superior downstream task performance.
This paper proposes a best-first search algorithm utilizing delayed partial expansions to explicitly treat control parameters as decision points within infinite domains, offering a complete and competitive alternative to existing constraint-based approaches for automated planning.
This paper introduces "Answer-Then-Check," a novel safety alignment method that enhances LLM robustness against jailbreak attacks by training models to generate direct answers internally and then critically evaluate their safety before responding, achieving superior protection with reduced over-refusal while maintaining general reasoning capabilities through the newly constructed 80K-sample ReSA dataset.
This paper addresses the inconsistency and structural biases in existing latency metrics for simultaneous speech-to-text translation by introducing a comprehensive meta-evaluation, proposing new metrics (YAAL and LongYAAL) and a resegmentation tool (SoftSegmenter), and implementing these solutions within the OmniSTEval toolkit to enable more reliable system assessments.
The paper introduces LikePhys, a training-free evaluation method using likelihood preferences to assess intuitive physics understanding in video diffusion models, demonstrating that current models show improving capabilities in physical reasoning as they scale despite challenges with complex dynamics.
Phys2Real is a real-to-sim-to-real reinforcement learning framework that enhances sim-to-real transfer for precise robotic manipulation by fusing vision-language model-inferred physical parameter priors with online interactive adaptation through uncertainty-aware ensemble estimation.