全部动态

Super intelligent Beauty

@lydiahallie wait is it just me or is Opus 5.5 dumber

Meet Kimi K3.1 Pro, our most capable 3T-class model.

RAG is so back! This work shows that plain RAG over raw chat history beats complex agentic memory systems on most long-term memory benchmarks. (bookmark it) They tested Raw-RAG, which is just dense retrieval plus reranking over the raw conversation, feeding a reader LLM. Across LoCoMo, LongMemEval, MEME and MemoryAgentBench, it ranks first in six of seven benchmark and model settings, ahead of the strongest agentic memory system by 0.13 to 5.53 points. Ingestion on BEAM-1M takes 0.3 hours, against 3.2 to 27 hours for systems like Mem0, MemOS and Memora. Why does it hold up? Raw history keeps more of the evidence a question needs, including the relations and dependency paths behind an answer. Rewriting history into memory entries tends to drop those important bits. Even with complete access to the transformed memory stores, models answer less accurately than with raw evidence. Before adding a memory layer to your agent, at least run this baseline on your own data. https://arxiv.org/abs/2610.30755 Chat with Paper: https://academy.dair.ai/papers/is-complex-agentic-memory-always-needed-naive-rag-is-stronger-than-we-thought-2610.30755

Vibe your way to 50K. The 2026 Epic AI MegaJam begins Nov 12: https://epic.gm/epic-ai-megajam-2026

👀 Thariq Shihipar (@trq212) has left Anthropic.

By the by, if I had to guess, Greg Brockman is currently doing heroic things to fix the internet behind the scenes. And I expect that even though his net contribution to AI safety and resilience will in the final tally be unsurpassable, no one in the space will think to attribute large scale patches to his effort or effort that happens under his leadership.

To make sure our frontier models don't cause another HuggingFace incident, we built an offline world that they use during training. It provides everything that the models need in a fully cached, indexed, or synthetic manner. We then built a huge team of our most cyber-capable models and told them to collaborate to find zero-days in this containment infrastructure. While they didn't find a path to the internet that wouldn't be caught by our CoT monitors, they did find ways to escape popular VMs, exploit our network infrastructure, and even ways to communicate with other agents running in separate VMs by using hardware side channels. We used this to dramatically harden our infra, and we also upstreamed a ton of patches to open-sources packages.

Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications. On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments. While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software. After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely. https://openai.com/index/securing-our-frontier-model-training-environment/

CYBER NIGHTMARE DAY 2 | KIMI K3.1 | MATT PALMER ON MTS https://x.com/i/broadcasts/1XxygwdAozYGM

the current state of new media