LLMs & Foundation Models· 2026-11-09
By Mr.Xu2 views0 discussions
Geographic Prompts Shift LLM Kidney Screening Results
A study on zero-shot chronic kidney disease (CKD) screening kept patient clinical inputs u…
ZICQ AI News
ZICQ curates daily AI news and deep dives on models, open-source infrastructure, and developer field reports.
Submit an articleLLMs & Foundation Models
View all (3163) →LLMs & Foundation Models· 2026-11-09
By Mr.Xu1 views0 discussions
vLLM Overload Study Shows Cancellation Alone Fails to Contain Retry Storms
A controlled study of vLLM under overload reveals that client-side retries can keep an LLM…
LLMs & Foundation Models· 2026-11-09
By Mr.Xu1 views0 discussions
ChainFlow Rebuilds Speculative Decoding with Two-Step Flow Matching
ChainFlow introduces a speculative-decoding drafter conditioned on the target model's hidd…
LLMs & Foundation Models· 2026-11-09
By Mr.Xu0 views0 discussions
Entropy Crossings in Recursive Self-Training Reveal Uneven Vocabulary Collapse
A study of nine generations of recursive self-training with GPT-2 reveals that Rényi entro…
LLMs & Foundation Models· 2026-11-09
By Mr.Xu0 views0 discussions
Biomedical LLM Fine-Tuning Audit Finds Erosion of Knowledge-Based Abstention
A paired audit evaluates 25 biomedical domain-fine-tuned language models against 11 corres…
Agentic
View all (2351) →Agentic· 2026-11-09
By Mr.Xu1 views0 discussions
Debugger Agents Are Not Necessarily Better: pass_X@k Exposes a Tool- Complementarity Illusion
This study isolates the effect of the debugging interface by keeping the coding-agent scaf…
Agentic· 2026-11-09
By Mr.Xu1 views0 discussions
Forge Decision Context Teaches AI Agents to Remember Corrections Across Sessions
Forge Decision Context (forge_dc) introduces persistent behavioral learning for LLM agents…
Agentic· 2026-11-09
By Mr.Xu1 views0 discussions
NormGym Tests Whether AI Agents Can Learn Hidden Social Norms
NormGym is an interactive benchmark for evaluating whether agents can acquire previously u…
Agentic· 2026-11-09
By Mr.Xu1 views0 discussions
Learning Least-Privilege Subagent Authorization from Agent Skills
This paper turns the procedural knowledge encoded in Agent Skills—operations, tools, and f…
Agentic· 2026-11-06
By Mr.Xu4 views0 discussions
Closed-Loop Agent Evaluation: Layered Measurement and Self-Calibration
This paper proposes closed-loop evaluation for deployed AI agents, treating evaluation as …
Open Source AI
View all (215) →Open Source AI· 2026-11-06
By Mr.Xu1 views0 discussions
NVIDIA Connects 6G Simulation, AI Training and Over-the-Air Testing with Open-Source Tools
NVIDIA's open-source AI Aerial, Sionna, and CUDA-accelerated RAN stack now connects severa…
Open Source AI· 2026-11-05
By Mr.Xu3 views0 discussions
NVIDIA Open-Sources Cosmos 3 Models for Interactive World Simulation
NVIDIA has released the Cosmos 3 open models through Hugging Face for interactive world si…
Open Source AI· 2026-11-04
By Mr.Xu1 views0 discussions
NetEase Youdao Open-Sources Confucius4-R2T2 Stable-Prefix Streaming ASR
NetEase Youdao has released and open-sourced Confucius4-R2T2, a streaming automatic speech…
Open Source AI· 2026-10-28
By Mr.Xu4 views0 discussions
Akai Embraces Open Source with MPC Sample Key, Inviting Community Modding
Akai is set to release the MPC Sample Key, an open-source hardware device aimed at music c…
Open Source AI· 2026-10-28
By Mr.Xu6 views0 discussions
Hugging Face Open-Sources Skill Dreaming: Optimizing Repository Skills via Pull Request Streams
Hugging Face has introduced the Skill Dreaming framework, designed to optimize repository-…
Compute & Architecture
View all (73) →Compute & Architecture· 2026-11-09
By Mr.Xu2 views0 discussions
NVIDIA Releases Dynamo-Optimized Recipes for Four Korean Open Models
NVIDIA has released Dynamo-optimized deployment recipes for A.X K2, K-EXAONE 2.0, Solar Op…
Compute & Architecture· 2026-11-06
By Mr.Xu4 views0 discussions
FlexFloor Elastically Releases MoE Training Workers, Boosting RL Rollout Throughput by 52.74%
FlexFloor addresses idle worker time in reinforcement-learning post-training for mixture-o…
Compute & Architecture· 2026-11-06
By Mr.Xu6 views0 discussions
LowRAM Compresses KV Cache with Low-Rank Keys and Amortized Values
LowRAM introduces an asymmetric, post-training pipeline for compressing the KV cache used …
Compute & Architecture· 2026-11-05
By Mr.Xu3 views0 discussions
NVIDIA Vera CPU Targets Agent Execution with High Throughput and Power Efficiency
NVIDIA is positioning the Vera CPU as the execution tier for AI agents, covering informati…
Compute & Architecture· 2026-11-05
By Mr.Xu6 views0 discussions
NVIDIA Tackles the AI Memory Wall with Full-Stack Codesign
NVIDIA is addressing the memory pressure created by larger models, longer contexts, and ag…
Research & Papers
View all (1797) →Research & Papers· 2026-11-09
By Mr.Xu0 views0 discussions
LLM Agents for Molecular Optimization Mostly Exploit Retrieval, Not Reasoning
An ablation study of low-budget molecular optimization benchmarks finds that much of the r…
Research & Papers· 2026-11-09
By Mr.Xu0 views0 discussions
Topology-Aware Optimization Cuts ImageNet Training Time by 14.1%
IMSLab Research introduces Topological-Adaptive Optimization, including Top-SGD and Top-Ad…
Research & Papers· 2026-11-09
By Mr.Xu1 views0 discussions
AttnRes Reweights Transformer Depth per Node for Long-Range Graph Reasoning
AttnRes replaces the fixed unit-weight accumulation used in Graph Transformer residual str…
Research & Papers· 2026-11-09
By Mr.Xu0 views0 discussions
GNN-Based Generative Models Reconstruct Missing Sensor Fields for Safety-Critical Digital Twins
Researchers at Oregon State University developed a sparse-sensor reconstruction framework …
Research & Papers· 2026-11-09
By Mr.Xu1 views0 discussions
QMAC Redefines Instruction-Tuning Data Selection with Multiband Distribution Matching
QMAC frames instruction-tuning data selection for large language models and large vision-l…
Industry & Trends
View all (632) →Industry & Trends· 2026-11-09
By Mr.Xu1 views0 discussions
NVIDIA Deepens Korea's Sovereign AI Push with NVLink Fusion
NVIDIA and South Korea’s Ministry of Science and ICT have expanded a collaboration spannin…
Industry & Trends· 2026-11-09
By Mr.Xu1 views0 discussions
NVIDIA and Korean Industry Leaders Bring Physical AI into Production
NVIDIA is expanding physical-AI deployments with Samsung Electronics, Kakao Mobility, and …
Industry & Trends· 2026-11-07
By Mr.Xu1 views0 discussions
NVIDIA and U.S. Commerce Face Scrutiny Over China H200 Resale Vetting
A New York Times investigation alleges that New Matrix, a Chinese entity approved to handl…
Industry & Trends· 2026-11-07
By Mr.Xu1 views0 discussions
AI Reshapes Cyber Defense as Cheap Agents Shrink the Response Window
In a Decoder interview, Cognition CISO Alex Stamos argues that open-weight models and comm…
Industry & Trends· 2026-11-06
By Mr.Xu1 views0 discussions
OpenAI Reportedly Raises Additional $11 Billion at $1.25 Trillion Valuation
OpenAI is reportedly pursuing an additional $11 billion in financing at a valuation of app…
大语言模型
大语言模型· 2026-10-09
By Mr.Xu9 views0 discussions
Hugging Face Releases PreSTO: A New Method for Accelerating LLM Inference
Hugging Face has introduced PreSTO, a novel method for accelerating the inference process …
Safety & Privacy
Safety & Privacy· 2026-08-10
By Mr.Xu11 views0 discussions
Checkpoint Research Exposes Potential Security Risks in Cloudflare Code Mode and Workers
Checkpoint Research has published a study on Cloudflare's Code Mode and Workers, highlight…