Open Omni Hits Flagship Scale, Self-Judge Breaks, Reasoning Leaks Forgotten Facts

Also Worth Noting

• Qwen3.5-Omni Technical Report score 4
  入选 HF Daily Papers; 关键词(4): latency, MoE, coding, reasoning
• Learning Uncertainty from Sequential Internal Dispersion in Large Language Models score 4
  关键词(1): deployment; 顶会接收: ACL
• Concept-wise Attention for Fine-grained Concept Bottleneck Models score 4
  关键词(3): fine-tuning, pre-training, vision-language; 顶会接收: CVPR
• TTL: Test-time Textual Learning for OOD Detection with Pretrained Vision-Language Models score 4
  关键词(1): vision-language; 顶会接收: CVPR
• Beyond Text Prompts: Precise Concept Erasure through Text-Image Collaboration score 4
  关键词(1): text-to-image; 顶会接收: CVPR
• CiPO: Counterfactual Unlearning for Large Reasoning Models through Iterative Preference Optimization score 4
  关键词(1): reasoning; 顶会接收: ACL
• Neural Gabor Splatting: Enhanced Gaussian Splatting with Neural Gabor for High-frequency Surface Reconstruction score 4
  关键词(3): lightweight, pruning, real-time; 顶会接收: CVPR
• MARCH: Multi-Agent Radiology Clinical Hierarchy for CT Report Generation score 4
  关键词(2): retrieval-augmented, vision-language; 顶会接收: ACL
• VEFX-Bench: A Holistic Benchmark for Generic Video Editing and Visual Effects score 4
  入选 HF Daily Papers; 关键词(2): vision-language, open-source
• FineCog-Nav: Integrating Fine-grained Cognitive Modules for Zero-shot Multimodal UAV Navigation score 4
  关键词(2): reasoning, vision-language; 顶会接收: CVPR
• The World Leaks the Future: Harness Evolution for Future Prediction Agents score 3
  机构: Tsinghua
• CoEvolve: Training LLM Agents via Agent-Data Mutual Evolution score 3
  顶会接收: ACL
• Exploring the Capability Boundaries of LLMs in Mastering of Chinese Chouxiang Language score 3
  顶会接收: ACL
• How Hypocritical Is Your LLM judge? Listener-Speaker Asymmetries in the Pragmatic Competence of Large Language Models score 3
  顶会接收: ACL
• Elucidating the SNR-t Bias of Diffusion Probabilistic Models score 3
  顶会接收: CVPR