A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Yizhao Gao, Shu-Yu Guo, Shijie Cao, et al.
SeerAttention-R adapts sparse attention for long-context reasoning, achieving strong AIME benchmark results with 4K token budgets and large sparse blocks.
Unknown
This paper introduces Llmc, a versatile toolkit for benchmarking LLM quantization methods, enabling standardized evaluation and comparison.
Unknown
This survey introduces Large Multimodal Reasoning Models (LMRMs) and proposes the concept of native LMRMs (N-LMRMs) that integrate perception, reasoning, and planning.
Songyue Han, Mingyu Wang, Jialong Zhang, et al.
A comprehensive survey of LLMs covering architectures, key technologies, interdisciplinary integrations, optimization, applications, and challenges.
Xianfu Cheng, Shiwei Zhang, Jiyu Zhao, et al.
FinanceComplexQA is a benchmark with 2,026 deep research tasks on 1,009 financial documents for evaluating agentic reasoning in complex financial QA.
Baohao Liao, Hanze Dong, Christof Monz, et al.
ReOPD enables efficient multi-turn on-policy distillation for LLM agents by reusing teacher trajectories as replayed prefixes, avoiding costly environment interactions.
Xiao Yu, Baolin Peng, Ruize Xu, et al.
OpenForgeRL is an open-source framework that enables end-to-end training of harness-based AI agents in any environment by decoupling training and inference via a lightweight proxy and Kubernetes orchestrator.
Wenhao Li, Xueying Jiang, Quanhao Qian, et al.
VLM-IE3D enhances 3D spatial awareness of vision-language models by fusing implicit and explicit 3D geometries learned from RGB videos.
Xu Wang, Kaixiang Yao, Miao Pan, et al.
Proposes ProVisE, a benchmark-agnostic framework to evaluate spatial cognition in image-generation models by parsing pixel-space outputs into structured predictions, and introduces SpatialGen-Bench for unified comparison with text-output VLMs.
Unknown
This paper investigates why smaller open-source long-context language models underperform proprietary ones in multi-hop reasoning with attributions.
Unknown
This paper identifies and addresses the reasoning bottleneck in Graph-RAG systems for multi-hop QA through structured prompting and context compression.
Unknown
This survey reviews reasoning capabilities in foundation models, covering concepts, methodologies, and future outlook.