AI 速递 2026-06-12
生成时间：2026/6/12 10:14:50（UTC: 2026-06-12T02:14:50.395Z）
Redesign Mixture-of-Experts Routers with Manifold Power Iteration
Section titled “Redesign Mixture-of-Experts Routers with Manifold Power Iteration”👍 76 · arXiv
Router is the cornerstone component to the Mixture-of-Experts models. Serving as expert proxies, the rows of the router matrix compute their similarity to the MoE inputs to determine which subset of e…
Toward Generalist Autonomous Research via Hypothesis-Tree Refinement
Section titled “Toward Generalist Autonomous Research via Hypothesis-Tree Refinement”👍 71 · arXiv
Scientific progress depends on a repeated loop of exploration, experimentation, and abstraction. Researchers test candidate directions, interpret the evidence, and carry the resulting lessons into lat…
Claw-SWE-Bench: A Benchmark for Evaluating OpenClaw-style Agent Harnesses on Coding Tasks
Section titled “Claw-SWE-Bench: A Benchmark for Evaluating OpenClaw-style Agent Harnesses on Coding Tasks”👍 56 · arXiv
General-purpose agents such as OpenClaw are increasingly used as autonomous tool users, but their coding ability is difficult to measure under SWE-bench: a generic agent does not by itself satisfy the…
Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application
Section titled “Agentic Environment Engineering for Large Language Models: A Survey of Environment Modeling, Synthesis, Evaluation, and Application”👍 56 · arXiv
Environments serve as interactive systems for large language model (LLM) based agents across diverse scenarios and play a crucial role in driving the continual evolution of model capabilities. Despite…
Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
Section titled “Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions”👍 52 · arXiv
Reward models are central to text-to-image post-training, but visual preference is subjective and better represented as a distribution over rubric scores than as a deterministic scalar. Existing scala…
OpenClaw v2026.6.6-beta.1
Section titled “OpenClaw v2026.6.6-beta.1”2026.6.6
Section titled “2026.6.6”Highlights
Section titled “Highlights”- Security boundaries are substantially tighter across transcripts, sandbox binds, host environment inheritance, MCP stdio, Codex HTTP access, native search policy, elevat…
链接：https://github.com/openclaw/openclaw/releases/tag/v2026.6.6-beta.1
LangChain langchain-model-profiles==0.0.6
Section titled “LangChain langchain-model-profiles==0.0.6”Changes since langchain-model-profiles==0.0.5
release(model-profiles): 0.0.6 (#38057) feat(standard-tests): validate tool call chunks during streaming (#34707) hotfix(core): bump lockfile(s) (#38032)…
链接：https://github.com/langchain-ai/langchain/releases/tag/langchain-model-profiles%3D%3D0.0.6
CrewAI 1.14.7
Section titled “CrewAI 1.14.7”What’s Changed
Section titled “What’s Changed”Features
Section titled “Features”- Add pluggable default backends for memory, knowledge, rag, and flow.
- Surface real finish_reason, sampling params, and response.id on LLM events.
- Type DSL triggers…
链接：https://github.com/crewAIInc/crewAI/releases/tag/1.14.7
OpenAI Codex CLI rust-v0.140.0-alpha.13
Section titled “OpenAI Codex CLI rust-v0.140.0-alpha.13”Release 0.140.0-alpha.13
…
链接：https://github.com/openai/codex/releases/tag/rust-v0.140.0-alpha.13
Theker just raised $85M to build the factory robot that doesn’t specialize in anything
Section titled “Theker just raised $85M to build the factory robot that doesn’t specialize in anything”Unlike humanoid robots designed around a fixed form — think Boston Dynamics — Theker’s machines are built to be reconfigured.
Jeff Bezos’s Prometheus raises $12B to build an ‘artificial general engineer’ for the physical world
Section titled “Jeff Bezos’s Prometheus raises $12B to build an ‘artificial general engineer’ for the physical world”The new round values the physical AI startup that aims to automate heavy engineering and drug design at $41 billion.
SpaceX officially prices shares at $135 in the largest IPO ever
Section titled “SpaceX officially prices shares at $135 in the largest IPO ever”Wits its official share pricing announcement, SpaceX’s IPO has begun.
SpaceX SPV investors won’t know their true holdings until post-IPO lock-ups lift
Section titled “SpaceX SPV investors won’t know their true holdings until post-IPO lock-ups lift”After SpaceX makes its public debut, lower-tier SPV investors face hidden fees, lengthy payout delays, and the risk of outright fraud.
Deezer’s new tool can identify AI music from Spotify, Apple Music, and others
Section titled “Deezer’s new tool can identify AI music from Spotify, Apple Music, and others”Deezer introduced a tool that scans playlists from Spotify, Apple Music, and other platforms to identify AI music.
Pool’s new app turns your screenshots into something useful
Section titled “Pool’s new app turns your screenshots into something useful”Pool’s new app automatically sorts screenshots into personalized collections, tracks down the original links behind saved content, and helps you rediscover products, recipes, travel ideas, and other things you meant to revisit.
DoorDash’s new AI chatbot lets you order with prompts and photos
Section titled “DoorDash’s new AI chatbot lets you order with prompts and photos”The new chatbot, called Ask DoorDash, allows users to search the app for what they’re looking for in their own words instead of having to scroll through restaurants and stores to build a cart.
Opendoor’s India exit is fueling a bigger conversation about AI and outsourcing
Section titled “Opendoor’s India exit is fueling a bigger conversation about AI and outsourcing”The decision comes as India emerges as the world’s largest GCC market.