AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
bluesky
bluesky-ai
12dMeta's new Ember model for agents 🤖 Meta released Muse Ember, a light, cost-efficient, multimodal AI model at 10 cents per million input …Meta's new Ember model for agents 🤖 Meta released Muse Ember, a light, cost-efficient, multimodal AI model at 10 cents per million i…
↗Not briefed
bluesky
bluesky-ai
12dClaude's user base has nearly doubled to 30 million this year, as more people turn to it for everyday tasks. But popular doesn't mean risk…Claude's user base has nearly doubled to 30 million this year, as more people turn to it for everyday tasks. But popular doesn't mean…
↗Not briefed
bluesky
bluesky-ai
12dAnthropic has unveiled new versions of its powerful Mythos and Fable AI modelsAnthropic has unveiled new versions of its powerful Mythos and Fable AI models
↗Not briefed
hn
hn-ai
12dFable 5.2 Available in Claude DesktopNew popup on my Claude app - Fable 5.2 is available for use right now, with a free Fable reset usable by March 8th, 2027.
↗Not briefed
hn
hn-ai
12dShow HN: Attentry – a local checker that reviews each edit Claude Code makes—
↗Not briefed
bluesky
bluesky-ai
12d#BREAKING Anthropic launches powerful new models ahead of Wall Street debut#BREAKING Anthropic launches powerful new models ahead of Wall Street debut
↗Not briefed
bluesky
bluesky-ai
12dWhat has electricity, nuclear forces, *and* antimatter? Why, it's the new web demo for Agent Higgs 2 of course! A particle physics puzzle…What has electricity, nuclear forces, *and* antimatter? Why, it's the new web demo for Agent Higgs 2 of course! A particle physics p…
↗Not briefed
hn
hn-ai
12dFable 5.2 Out—
↗Not briefed
arxiv
arxiv-ai
12dASPIRE: An Agentic System-Identification Platform for Iterative Reasoning and ExplorationSystem identification requires expert decisions on the identification method, incorporation of prior physical knowledge, and model val…
↗Not briefed
bluesky
bluesky-ai
12dClaude Fable 5.2 is designed to take on big technical challenges, and it’s available today. www.theverge.com/ai-artificia...Claude Fable 5.2 is designed to take on big technical challenges, and it’s available today. www.theverge.com/ai-artificia...
↗Not briefed
bluesky
bluesky-ai
12dAnthropic releases Mythos 5.2 and Fable 5.2, saying they are "best for the most ambitious, open-ended technical work", for $8/1M input and …Anthropic releases Mythos 5.2 and Fable 5.2, saying they are "best for the most ambitious, open-ended technical work", for $8/1M input…
↗Not briefed
arxiv
arxiv-ai
12dBuilding World Models from Intrinsic MotivationIn the absence of extrinsic rewards or task supervision, human behavior is guided by intrinsic motivation---neural signals that drive…
↗Not briefed
rss
techmeme
12dAnthropic releases Mythos 5.2 and Fable 5.2, saying they are "best for the most ambitious, open-ended technical work", for $8/1M input and $40/1M output tokens (Anthropic)Anthropic: Anthropic releases Mythos 5.2 and Fable 5.2, saying they are “best for the most ambitious, open-ended technical work”, for…
↗Not briefed
hn
hn-ai
12dSTM32 Memory Protection in Practice—
↗Not briefed
bluesky
bluesky-ai
12dHello developers! 🎮 Dive into the gozinaki API for TinCLE and embed lightning-fast on-device LLM intelligence directly into your game loop…Hello developers! 🎮 Dive into the gozinaki API for TinCLE and embed lightning-fast on-device LLM intelligence directly into your game…
↗Not briefed
arxiv
arxiv-ai
12dCatch It Where It Happens: Real Time Error Correction During LLM ReasoningLarge language models (LLMs) are increasingly trusted with mathematical, scientific and engineering work, yet their mistakes are often…
↗Not briefed
bluesky
bluesky-ai
12dSeriously, who among us haven't slept with a foreign spy? Let they who throw stones in glass houses, etc, etc... In short, the LLM grift …Seriously, who among us haven't slept with a foreign spy? Let they who throw stones in glass houses, etc, etc... In short, the LLM g…
↗Not briefed
arxiv
arxiv-ai
12dOne Temperature to Rule Them All?Inference-time sampling is a central mechanism for improving large language models, with temperature controlling the tradeoff between…
↗Not briefed
rss
theverge-ai
12dAnthropic launches new Claude Fable and Mythos 5.2 modelsAnthropic is announcing two new models today: Claude Fable 5.2 and Claude Mythos 5.2, which it says are designed to take on "daunting"…
↗Not briefed
rss
openai-news
12dChatGPT Enterprise now available in AWS Marketplace🔧 ChatGPTChatGPT Enterprise is now available through AWS Marketplace, helping organizations simplify procurement through existing AWS purchasin…
↗Not briefed
arxiv
arxiv-ai
12dGuiding Data Allocation for Robust Subpopulation GeneralizationMachine learning models often achieve high average accuracy in the training data while performing poorly on test data due to subpopula…
↗Not briefed
arxiv
arxiv-ai
12dIC-GNN: Modeling Individualized Brain Community Organization for Disease PredictionBrain networks exhibit modular organization, where community structures vary substantially across individuals and diseases. Understand…
↗Not briefed
arxiv
arxiv-ai
12dKolmogorov-Chapman Maps: Trajectory-Free Learning for Simulating Molecular DynamicsStochastic differential equations (SDEs) describe stochastic processes across a wide range of scientific domains, including physics, b…
↗Not briefed
hn
hn-ai
12dShow HN: Search 17 years of "What are you working on?"Inspired by Karpathy's HackerTube, I wondered what was buried in "What are you working on?" thread comments. I wanted t…
↗Not briefed
arxiv
arxiv-ai
12dINTENT: Out-of-Sample, Out-of-Time, and Cold-Start Evaluation on User Event SequencesBanks now train foundation models on ordered sequences of customer events such as transactions, app sessions, and authorizations. Depl…
↗Not briefed
hn
hn-ai
12dTesting Claude on real-world electronics design—
↗Not briefed
arxiv
arxiv-ai
12dAdaptive Quantization Preserves Fast Rates for TD Learning over Rate-Limited ChannelsMotivated by reinforcement learning subject to communication constraints, we study a policy evaluation problem with linear function ap…
↗Not briefed
bluesky
bluesky-ai
12dThe problem with automated AI detectors is the problem with AI. Writing and reading are both complex skills and, while it is impressive how…The problem with automated AI detectors is the problem with AI. Writing and reading are both complex skills and, while it is impressiv…
↗Not briefed
hn
hn-ai
12dClaude Fable 5.2—
↗Not briefed
hn
hn-ai
12dClaude Code memories that save the fix, not just a note—
↗Not briefed
arxiv
arxiv-ai
12dHiVe-Mem: Interleaving Horizontal Memory Expansion and Vertical Evidence Composition for Multimodal AgentsAs multimodal agents accumulate increasingly long interaction histories, effectively retrieving and utilizing past information becomes…
↗Not briefed
bluesky
bluesky-ai
12dEven in the context of 'great AI tech' like Claude, it's mostly just a way to generate a lot of code that you do not grasp, along with some…Even in the context of 'great AI tech' like Claude, it's mostly just a way to generate a lot of code that you do not grasp, along with…
↗Not briefed
hn
hn-ai
12dClaude Fable 5.2 and Mythos 5.2—
↗Not briefed
arxiv
arxiv-ai
12dVendi Anomaly Scores for Efficient and Accurate Anomaly DetectionAnomaly detection (AD) is a critical for trustworthy machine learning and scientific discovery. Existing methods typically define anom…
↗Not briefed
arxiv
arxiv-ai
12dWhere Does Alignment Emerge and What Does It Mean? A Closer Look at the Platonic Representation Hypothesis🔧 ModalThe Platonic Representation Hypothesis (PRH) suggests that neural networks are converging toward similar representation spaces. As lan…
↗Not briefed
arxiv
arxiv-ai
12dShortcuts and Boundaries: Verifiable RL Misalignment and Contextual Safety LearningReinforcement learning (RL) can improve an agent's task capability while reinforcing behavior that violates contextual execution requi…
↗Not briefed
hn
hn-ai
12dShow HN: MergeAtlas – A better way to review PRsHey HN community, I built MergeAtlas - a better way to review PRs when bulk of the code is written by AI agents. It is a skill for you…
↗Not briefed
hn
hn-ai
12dSynaptics sends cease-and-desist order to open-source DisplayLink hobbyist—
↗Not briefed
arxiv
arxiv-ai
12dIs Persona Generalization Predictable?Persona generalization occurs when fine-tuning a language model in one domain induces behavioral changes in others -- for example, the…
↗Not briefed
rss
techmeme
12dMavria, which aims to build modular data centers that can be shipped by trucks, raised $130M across pre-seed and seed rounds at a $1B valuation led by Paradigm (Jennifer Hiller/Wall Street Journal)Jennifer Hiller / Wall Street Journal: Mavria, which aims to build modular data centers that can be shipped by trucks, raised $130M ac…
↗Not briefed
rss
the-decoder
12dBytedance's Seedream 5.5 promises pixel-stable image editing across 100 rounds without color driftBytedance Seed has introduced Seedream 5.5, an image model built for professional production workflows. It supports stable multi-pass…
↗Not briefed
bluesky
bluesky-ai
12darstechnica.com/ai/2026/11/c...arstechnica.com/ai/2026/11/c...
↗Not briefed
bluesky
bluesky-ai
12dFive protestors arrested after taking over the lobby at Anthropic. This is an activist from @climatedefenders.bsky.social . Othe groups inc…Five protestors arrested after taking over the lobby at Anthropic. This is an activist from @climatedefenders.bsky.social . Othe group…
↗Not briefed
hn
hn-ai
12dE-means: 10x faster k-means at equal loss—
↗Not briefed
arxiv
arxiv-ai
12dWorkflowArena: Benchmarking Executable Workflow Orchestration in LLM-Based Multi-Agent SystemsLarge language model agents are increasingly deployed as multi-agent systems rather than as a standalone model operating within a sing…
↗Not briefed
arxiv
arxiv-ai
12dWhat kills $v$-prediction? A Patch-wise PCA Perspective on Pixel-Space Flow MatchingLatent Diffusion Models dominate generative modeling but face two-stage training complexity and the inherent reconstruction limits of…
↗Not briefed
bluesky
bluesky-ai
12dNVIDIA'S JENSEN HUANG NO LONGER SET TO ATTEND MONDAY ANTHROPIC CEO FORUM - BUSINESS INSIDER.NVIDIA'S JENSEN HUANG NO LONGER SET TO ATTEND MONDAY ANTHROPIC CEO FORUM - BUSINESS INSIDER.
↗Not briefed
hn
hn-ai
12dPolinRider: North Korea-linked worm forges developers' commits in 7,609 repos—
↗Not briefed
hn
hn-ai
12dChaCha AVX-512 Optimization for Linux 7.5 Yields Up to 99% Improvement AMD Zen5—
↗Not briefed
hn
hn-ai
12dShow HN: MeowDisplay Is iPhone Mirroring, but in ReverseHey HN and Cats, thank you for checking this out!<p>Github: <a href="https:&#x2F;&#x2F;github.com&#x2F;raiseCatError&#x2F;meowdisplay"…
↗Not briefed