ExploreRepair agents patch vulnerabilities without full bug details
On PatchEval it scores 39.6% with GPT-5.6-Luna and 71.6% with DeepSeek-V4.1-Flash, above EvoRepair by 16.0 and 25.4 points.
46
- Mentions · 7d1
- Mentions · 30d10
- All mentions15
- First seenSep 10, 2026
- Last mentioned1d ago
Timeline
15 articles · newest firstLow-bit n-gram embeddings cut table memory with little loss
Four-bit tables saved about 75% of table memory on models trained from scratch while keeping over 99% of the validation-loss gain.
42
Storm-0940 used ARTEX and AI coding agents against many targets
Microsoft says the group tried more than 2,250 tenants at about 2,900 sign-ins a minute.
74
Zscaler details AI setup behind a global hacking campaign
Jump host 38.244.50.120 routed connections to targets while ARTEX requested DeepSeek through an unverified gateway.
74
US orders federal civilian agencies to purge DeepSeek within 60 days
The ban covers Hangzhou DeepSeek technology and third-party tools built on it, with written CISA exemptions possible.
78
DeepSeek open-sources DeepInfer, inference stack for V4.1-Flash
The feed contained only the headline; license, hardware targets, and release notes were not in the text.
48
Microsoft and NVIDIA say open-weight AI shifts security burden to enterprises
Guidance covers 30 risk types, unsigned model files, and backdoors that behavioral tests may miss.
48
GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece
OpenAI's GPT-6-Astra solved 49 of 50 routine theoretical-physics problems at about $2.40 each.
58
BizBench tests long-horizon payment agents on constraint-valid execution
Claude Opus 5 and GPT-6 Sol reach only 64.0% and 61.3% strict success on 50 payment tasks.
48
A Closer Look at Agentic BBO: Benchmarking LLM Agents for Black-Box Optimization
AgenticBBO-Bench compares LLM agents on black-box optimization across five domains, with GPT-6 Astra on the cost-performance frontier.
50
The AI Race Just Got Awkward
A commentary says Western labs adopted DeepSeek KV-cache optimizations and cut cache-read prices.
34
DeepSeek-v4.1 Flash: Pushing the Limits of KV Cache Compression
DeepSeek-V4.1 Flash is a 552B-parameter multimodal MoE model with 1M-token context achieving 4x KV cache compression for long-horizon agent workloads.
74
DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression
DeepSeek releases V4.1-Flash, a 552B-parameter multimodal MoE with 1M-token context and 890-byte-per-token KV cache for cheaper agentic workloads; weights on Hugging Face.
84
DeepSeek AI Released DeepSeek-V4.1-Flash with 1M Context, FP4 KV Cache, and Cross-Layer Attention Reuse
DeepSeek released open-weight V4.1-Flash, a 552B MoE model with 1M context and FP4 KV cache, beating Opus-5 and GPT-5.6 Sol on agent benchmarks.
80
deepseek-ai/DeepSeek-V4.1-Flash — new model trending #28 on Hugging Face
DeepSeek releases DeepSeek-V4.1-Flash, a 552B-parameter multimodal MoE model with 1M-token context and KV cache cut to 890 bytes per token.
80
Appears with
- DeepSeekorganization · 6
- DeepSeekvendor · 5
- Claude Codeproduct · 3
- Hugging Faceorganization · 3
- Codexproduct · 3
- DeepSeek-V4-Flashai model · 3
- OpenAIorganization · 2
- Claude Opus 5ai model · 2
- OpenAIvendor · 2
- Anthropicvendor · 2
- DeepSeek-V4.1-Flashproduct · 2
- ARTEXmalware · 2
- GPT-6 Astraai model · 2
- Recorded Futureorganization · 2
- Anthropicorganization · 2
- deepseek-v4-proai model · 2
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.