-
Harness Engineering for Software Engineering via Modular Executable Dev-Primitives
score 6
入选 HF Daily Papers; HF 热度: 6 upvotes (+2); 关键词(1): reasoning
-
VeriFine: Scaling Verification for Self-Improvement in Embodied Reasoning
score 6
入选 HF Daily Papers; HF 热度: 5 upvotes (+2); 关键词(4): scaling, fine-tuning, reasoning, embodied
-
Evaluating human-AI workflows for field research in viticulture
score 4
机构: Cornell; 关键词(1): throughput
-
WASD: Wasserstein-based Knowledge Distillation for Large Language Models
score 4
关键词(3): distillation, code generation, reasoning; 顶会接收: NeurIPS
-
Cite What You Explore: Budget-Aware LLM Reasoning over Medical KGs with Verifiable Evidence
score 4
关键词(1): reasoning; 顶会接收: NeurIPS
-
OTel: Open Telco AI Datasets, Benchmarks, and Models
score 4
关键词(2): instruction tuning, post-training; 顶会接收: NeurIPS
-
Illusory Pattern Perception Drives Spurious Inference in Large Language Models
score 4
关键词(1): reasoning; 顶会接收: NeurIPS
-
ThinkFuse: Trajectory-Aware Test-Time Fusion for Small Reasoning Models
score 4
关键词(1): reasoning; 顶会接收: EMNLP
-
OpenWAM: An Open Framework for Composable World-Action Models
score 4
机构: Stanford; 关键词(1): pretraining
-
EmbodiedSmith: Scaling Embodied Data through Recursive Self-Improvement Flywheel in Simulation
score 7
入选 HF Daily Papers; HF 热度: 12 upvotes (+3); 关键词(4): scaling, pretraining, agentic, embodied
-
Performative Prediction with Selective Labels
score 4
关键词(1): deployment; 顶会接收: NeurIPS
-
Foresight-over-Graph: Reasoning Beyond Local Horizons for Knowledge Base Question Answering
score 4
关键词(2): pruning, reasoning; 顶会接收: NeurIPS
-
NeMo-DCR: Bit-Exact Delta-Compressed Refit for Scalable Agentic RL at Trillion-Parameter Scale
score 12
机构: NVIDIA; 入选 HF Daily Papers; HF 热度: 13 upvotes (+3); 有代码实现; 关键词(2): serving, agentic
-
SquidAgent: Parallelize Wisely, Coordinate Efficiently
score 4
关键词(2): latency, throughput; 顶会接收: NeurIPS
-
A Case Study in Assuring AI-Written Software
score 4
机构: Meta; 关键词(2): production, coding