#推理
Today 1 items
The DecoderAI score5959
The DecoderAI score6767 Mathematicians say OpenAI doesn't understand its own Navier-Stokes proof
Mathematicians dispute OpenAI's 166-page potential proof of the Navier-Stokes problem published in September, arguing that although it is formalized in Lean, it only superficially resembles the approaches of Diego Córdoba and Luis Martínez-Zoroa as OpenAI claimed; the CMI is reviewing it.
TechCrunch · AIAI score5454 Elorian’s first AI model wants to reason with images, not just about them
Palo Alto AI startup Elorian released its first model Glimpse on Friday, available as an API in research preview, with a chat interface limited to a waitlist. Glimpse can crop image regions, extract video frames and run code to take measurements during reasoning, which the company calls a visual scratch space.
The DecoderAI score4444 OpenAI's new Rosalind Academic Plan gives research labs up to 30 free seats for a year
OpenAI launched the Rosalind Academic Plan, currently open only to research labs at US universities; eligible labs can get up to 30 free seats on the Business Standard plan for up to twelve months, excluding API access.
Hugging Face BlogSelectedAI score7777 Welcome Beam: a 501B open model you can own, customize, and scale
Reflection has open-sourced its first model, Beam (reflectionai/Beam-1.0-501B-A23B), on Hugging Face: a 501B-parameter MoE model (23B activated per token) with 1M token context, released under the Apache 2.0 license and targeting coding and agent use cases.
Why it matters: The original provides architecture details, benchmark results, hardware requirements, and fine-tuning recipes, so readers can assess the feasibility of deploying and customizing Beam on their own hardware.
Hugging Face BlogAI score5555 Granite Switch 4.1: Chat, Decisions, and Speech from One Endpoint
IBM released the Granite-Switch-4.1-8B and -30B models, combining natural language responses, general and specialized decisions, and speech-to-text into a single managed endpoint via Decision aLoRA, Guardian aLoRA, and Granite-Speech-5.0-Turbo-CTC, with memory usage close to the base model.
The Verge · AIAI score7373 Meta and Apple team up to make Muse AI models run better on Macs
Meta released Muse Spark Open, an open-weights model it calls the strongest American open-weights model yet, with 816B total parameters, natively multimodal, excelling at long-horizon tasks, complex software engineering and multi-step reasoning. Meta partnered with Apple to optimize the model, which can run on desktop via the latest MLX framework on two Apple M5 Ultra Mac Studios; Apple also published a technical article on MLX, and executives from both companies exchanged public pleasantries on X.
Mistral AISelectedAI score6060 Introducing Timestral
Mistral has released Timestral, its first family of time series foundation models in four sizes — 5M, 28M, 123M and 209M parameters — with the largest model ranking first among pretrained zero-shot models on the GIFT-Eval, TIME and fev-bench public benchmarks, beating the previous SOTA with about two-thirds the parameters.
Why it matters: The article gives verifiable details on the four sizes, benchmark rankings and CPU inference latency, letting readers evaluate it for time series forecasting deployments.
Hugging Face BlogSelectedAI score6464 Apertus 1.5 Release Notes
Project Apertus has released Apertus 1.5, adding image and audio understanding and an optional thinking mode to its 8B and 70B models via continued pretraining, and extending the context window to 262,144 tokens — 4x that of Apertus 1.0.
Why it matters: The official release notes fully document the training pipeline, open data and intermediate checkpoints, letting readers trace how the multimodal and reasoning capabilities were built.
NVIDIA BlogAI score2929 The Nordics and NVIDIA Show What’s Possible When a Region Commits to Advanced Computing
The Nordics are building a leading European technology ecosystem with NVIDIA technology, expected to deploy 1 gigawatts of AI cloud compute by the end of 2027.
The DecoderAI score6464 OpenAI will reportedly ship at least one model this year but its flagship still needs alignment work
According to The Information, citing sources, OpenAI may release one or two new models before the end of the year, but the flagship model under development still needs alignment work and test infrastructure, and under an executive order it must be made available to the US government for evaluation up to 30 days in advance.
Google ResearchSelectedAI score7474 Analyzing technical documents with Deep Review
Google Research has publicly released Deep Review, an agent system that provides rigorous feedback and technical verification for complex technical documents. Formerly the Paper Assistant Tool (PAT), it partnered with four conferences — STOC 2026, ICML 2026, NeurIPS 2026 and ICLR 2027 — to deliver pre-submission feedback on over 16,000 opted-in manuscripts.
Why it matters: The article details real-world usage of over 16,000 manuscripts across four top conferences and the four-stage agent pipeline design, which can transfer to your own document review workflows.
Google ResearchAI score5050 Putting Earth AI to work: Geospatial agents for our planet's biggest challenges
Google has upgraded its Gemini-based Geospatial Reasoning agent into an experimental system capable of multi-step reasoning across planetary-scale data and autonomously training custom prediction models, and released a technical report introducing two new benchmarks, GeoKTools and GeoExpertQueries.
The Verge · AIAI score6262 Microsoft has a new machine learning tool to improve weather forecasts
Microsoft researchers released Probabilistic Bias Correction (PBC), a machine learning tool that improves the accuracy of two-to-six-week forecasts by correcting systematic biases in existing weather models. The paper was published in Nature and the tool is open source.
Simon WillisonAI score6060 llm-meta-ai 0.2
Meta AI released its new flagship model Muse Wonder (and Muse Wonder Full), which the author describes as Meta's first possible Fable-tier models; Artificial Analysis scores Muse Wonder (Max) as equivalent to GPT-6.1 Sol. Simon Willison updated his llm-meta-ai plugin to 0.2.
Google DeepMindSelectedAI score7878 Developing Gemini 4 Argon responsibly and safely
Google DeepMind released Gemini 4 Argon, focused on complex tasks like coding, finance, law, cybersecurity and science, and opened it first to cyber defenders through the Fairwind Program.
Why it matters: Google also details specific defenses against misuse, misalignment and prompt injection along with quantified results, letting readers see how frontier model safety is actually done.
Hugging Face BlogSelectedAI score6767 Run fast, private agents on-device with LFM2.5-24B-A2B
Liquid AI released LFM2.5-24B-A2B, a reasoning model with 24B total parameters and only 2B activated per token, aimed at fast private on-device agents, published with open weights on Hugging Face.
Why it matters: The release includes full benchmark comparisons, inference speed and a five-stage training pipeline, so readers can judge whether it fits local agent workflows.
Ars Technica · AISelectedAI score8080 Gemini 4 Argon begins rolling out in Gemini app and Antigravity, but only if you pay
Google's Gemini 4 Argon began rolling out today in the Gemini app, limited to AI Pro ($20/month) and AI Ultra (from $99/month) subscribers.
Why it matters: The article maps out Argon's availability across subscription tiers and the API along with limited-time pricing, helping developers gauge access costs.
Google DeepMindSelectedAI score8383 Gemini 4 Argon: Built to solve your hardest problems
Google DeepMind released Gemini 4 Argon, available worldwide to developers, enterprises, scientists and consumers via the Gemini API, AI Studio, Antigravity and the Gemini app, priced at $2 per million input tokens and $10 per million output tokens, with cached input tokens at a 95% discount.
Why it matters: The official release includes API pricing, engineering and science use cases and safety data, letting readers assess its real value for coding and agent scenarios.
The Verge · AISelectedAI score8080 Google finally releases Gemini 4 Argon
Google released the Gemini 4 Argon model, claiming frontier reasoning capabilities that beat GPT-6 Astra, Claude Fable 5.1 and Claude Opus 5.5 on benchmarks in coding, cybersecurity, science, finance and law.
Why it matters: The article covers Argon's pricing, subscription availability and internal use cases, letting readers compare its price gap and actual usability against competitors.
AWS Machine Learning BlogAI score3737 Mixture-of-Experts kernel engineering on Amazon SageMaker AI
The AWS team implemented kernel optimizations for Fused MoE models on Amazon SageMaker AI; the Swap-AB trick flips matrix multiplication before entering PTX assembly to fit the M=64 dimension requirement of NVIDIA Hopper's WGMMA instructions, delivering a 2–4% end-to-end throughput gain on H200, and has been upstreamed into vLLM 0.24.0 and later.
The DecoderSelectedAI score8181 Meta's new frontier model Muse Wonder Full breaks into the AI top tier
Meta released the frontier models Muse Wonder and Muse Wonder Full, which score 52 and 56 respectively on the Artificial Analysis Intelligence Index v4.3, up from 48 for the previous Muse Spark 1.3.
Why it matters: The benchmark data is fairly complete, letting readers compare Meta's new models with Claude, GPT, and Gemini on intelligence index, coding, and cost.
TechCrunch · AISelectedAI score7777 Meta releases its first frontier model, Muse Wonder
Meta Superintelligence Lab released Muse Wonder and Muse Wonder Full, Meta's first attempt at frontier models. According to Meta's model card testing, both models' benchmark scores are on par with Fable 5.1 and GPT-6 Astra.
Why it matters: The article gathers pricing, API availability, MemoryBench, and other key details, making it easy to compare Meta's first frontier models with competitors on the same footing.
Import AIAI score4949 Import AI 476: Creativity and RSI; worse-than-China open weight models; AI goes to space
AI startup Elorian and Scale AI released the visual intuition reasoning benchmark Humanity's Sixth Sense (HSS), containing 522 questions, 288 images and 234 short videos (17.6 hours of video in total).
The DecoderAI score7575 Google Deepmind's AlphaProof Nexus solved dozens of open math problems and actually got peer reviewed
Research published in Science by Google DeepMind shows its AlphaProof Nexus independently solved 9 of 353 open Erdős problems, two of which had stood for 56 years, proved 44 conjectures in the OEIS integer sequence database, solved an open algebraic geometry problem, and improved known bounds for min-max optimization.
The DecoderAI score5757 Microsoft's Decision-1 model enters the fast-growing AI decision model race
Microsoft launched Decision-1, a decision model for classification, evaluation, and routing decisions, built on Qwen3.5-9B.
The DecoderAI score4040 Google's Gemini 4 "Carbon" model reportedly feels like Anthropic's Opus 5.5 coding performance
According to Business Insider, Google is testing several Gemini 4 variants — Argon, Barium, and Carbon — with Carbon recently deployed on its internal coding platform Jetski, reportedly outperforming Argon on coding tasks; some employees likened it to Anthropic's strongest coding model, Opus 5.5.
Latent SpaceAI score4747 Why AlphaFold Didn't Solve Protein Folding — Pushmeet Kohli, Google DeepMind & Sal Candido, Biohub
Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido discuss why AlphaFold's breakthrough was only a starting point and protein structure prediction is far from solved.
Latent SpaceAI score4343 [AINews] not much happened today
OpenAI fired three safety researchers — Tomek Korbak, Mikita Balesni and Jasmine Wang — who say they were let go for 'putting safety above the company's short-term interests.'
TechCrunch · AIAI score6868 OpenAI’s math solutions aren’t meeting the field’s standards yet
OpenAI this week released hundreds of mathematical proofs claiming to solve major open problems, but several requirements in the AGMAI guidelines from Princeton's Institute for Advanced Study were unmet: of 719 submissions, only 10 included the model's chain of thought, only 42% of proofs went unformalized, and the required machine-readable metadata linking natural language to formal artifacts was not provided.
Latent SpaceAI score5353 Synthesis Superintelligence: from Semiconductors to Superconductors — Periodic Labs’ Liam Fedus and Ekin Dogus Cubuk
The Latent Space podcast interviews Periodic Labs co-founders Liam Fedus and Ekin Dogus Cubuk about their "synthesis superintelligence" vision: building reinforcement learning environments around physical labs where AI agents learn through an end-to-end cycle of predicting, synthesizing and characterizing materials, rather than relying only on internet data.
The DecoderSelectedAI score7676 Claude Haiku 5.5 arrives with massive price cuts proving the AI pricing arms race is far from over
Anthropic released Claude Haiku 5.5, cutting average prices by about 75% versus Haiku 4.5, with discounts of up to 90% for prompts under 100,000 tokens, and for the first time offering adjustable reasoning levels in a Haiku-tier model.
Why it matters: The article compiles Haiku 5.5's price cuts, key benchmark data and adjustable reasoning levels, allowing readers to compare its positioning against GPT-6 Luna and Sonnet 5.5.
Hugging Face BlogSelectedAI score7373 One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO
The NVIDIA team, building on Nemotron 3, used SFT, RL, and a generate-verify-improve reasoning system to score 30/42 at IMO 2026 (gold threshold 29), while Nemotron-3-Ultra-CC scored 535.4/600 in unofficial IOI 2026 testing (gold threshold 361.12).
Why it matters: The article lays out the complete fine-tuning and inference recipe from base model to gold medal, with open weights, data, and pipelines, so readers can reproduce this domain-specialization approach.
The DecoderSelectedAI score7777 OpenAI dumps 372 AI-generated math proofs on GitHub, telling the academic world to keep up
OpenAI published 372 new mathematical results generated by its internal frontier model on GitHub, covering significant algorithmic improvements and progress related to the Riemann hypothesis, with each result claimed to solve an open problem or make substantial progress.
Why it matters: The article examines how OpenAI released 372 AI-generated math proofs and the academic disagreement over it, helping readers understand how formal verification addresses the peer-review bottleneck.
Simon WillisonAI score3636 Quoting Jake Boggan
Jake Boggan, who once moved to Budapest to research graph theory, said he invested 24 years and thousands of hours, on and off, into the Barnette conjecture, and felt deeply conflicted upon learning it was reportedly proven in openai/math problem 180. He described the feeling as "as sad as hearing an ex-girlfriend suddenly died in a car crash," and guessed many others would feel equally mixed tonight.
The DecoderAI score6060 Researchers stretch LeCun's JEPA AI into a universal world model that works from physics to biology
A research team led by PhAI Labs together with Stanford, Oxford and Princeton has released JEPA-Anything, extending Yann LeCun's JEPA architecture into a general-purpose world model that works across seven domains.
The DecoderAI score6060 Reflection's Beam becomes the most capable open-weight model built outside China
Reflection announced its first free open model, Beam, with 501B total parameters and 23B activated per token (MoE), aimed at coding, logical reasoning and agentic tasks. According to Reflection, Beam matches GLM 5.2 on hard reasoning tasks with three to four times less compute, and comes close to Qwen3.8-Max on coding and agentic benchmarks, though its raw performance still trails stronger open models like Kimi K3.
OpenAI NewsSelectedAI score6666 Sharing AI progress in mathematics
OpenAI released new results on its internal frontier model solving open mathematical problems, and published Lean proof formalizations and research details on GitHub.
Why it matters: The article reports that OpenAI shared results from its internal frontier model on open math problems and released Lean formalized proofs and research details on GitHub.
Microsoft ResearchAI score4848 What AI gets wrong and what failure teaches us
In a Microsoft Research podcast, partner research manager Jennifer Neville speaks with applied scientist Chad Atalla about how evaluation drives AI systems past performance limits to meet user needs, and about the "unexpected failures" models show when tested beyond standard benchmarks. She shares practical advice on using existing AI systems and stresses the importance of carefully checking data when results don't match expectations.
TechCrunch · AIAI score5959 Reflection debuts Beam, an open-weight AI model to rival Chinese models at lower compute cost
US startup Reflection AI officially released Beam, its first frontier open-weight model, claiming it performs comparably to Z.ai's GLM-5.2 on advanced reasoning benchmarks while using 3-4x less inference compute.