News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
hn hn-ai | 29d | Ask HN: How bad Earth would have to be for terraforming Mars to make sense? | Ok, this might be more of a Reddit question or an xkcd’s “What if”, but let’s try here first.<p>When people talk about investing in te… | ↗Not briefed |
arxiv arxiv-ai | 29d | Enabling Token-wise Importance Sampling in Off-Policy RLVR with Proximal KL Regularization | Off-policy data in reinforcement learning with verifiable rewards (RLVR) can arise from asynchronous generation, rollout-training mism… | ↗Not briefed |
arxiv arxiv-ai | 29d | SP$^2$ec: Adaptive Self-Speculative Decoding for Vision-Language Models🔧 Modal | Speculative Decoding (SD) has demonstrated its efficacy in accelerating large language model inference, but for Vision-Language Models… | ↗Not briefed |
bluesky bluesky-ai | 29d | 🎨 Turn your ideas into 3D assets with AI! Hyper3D uses Rodin to generate detailed, textured 3D models from text prompts or images. ✅ Gen… | 🎨 Turn your ideas into 3D assets with AI! Hyper3D uses Rodin to generate detailed, textured 3D models from text prompts or images.… | ↗Not briefed |
bluesky bluesky-ai | 29d | Okay took a second but it synced tangled.org/pfrazee.com/... | Okay took a second but it synced tangled.org/pfrazee.com/... | ↗Not briefed |
arxiv arxiv-ai | 29d | Learning When to Transfer: Reinforcement Learning for Carbon-Footprint Modeling in New Manufacturing Domains | Manufacturers often need to characterize a new material, process, or product before enough operating data are available to train relia… | ↗Not briefed |
hn hn-ai | 29d | Show HN: The first open-weight model for general deterministic task | Hey HN! Today we’re releasing interfaze-1.1-lite, an open-weight model for tasks with a deterministic answer: OCR, STT with speaker di… | ↗Not briefed |
rss huggingface-blog | 29d | Run fast, private agents on-device with LFM2.5-24B-A2B | — | ↗Not briefed |
arxiv arxiv-ai | 29d | On the Monitorability of Natural Language Autoencoders: Identification vs Discrimination🔧 Modal | Natural Language Autoencoders (NLAs) are proposed as oversight tools for language models: an Activation Verbaliser (AV) writes an Engl… | ↗Not briefed |
arxiv arxiv-ai | 29d | Code-Switching Detection via Language-Selective Dimensions in Speech Foundation Models🔧 Whisper | Code-switching (CS), the alternation between languages within an utterance, is common in multilingual speech but underrepresented in p… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Problems – see where agents get stuck using your product | hi hn, i'm joey, the founder of problems.dev.<p>agents are increasingly becoming the most important customers of products. but w… | ↗Not briefed |
arxiv arxiv-ai | 29d | Learning Source Structure: Bayesian Prediction with Transformers | A transformer trained on sequences whose source parameters are drawn afresh each time can learn only what the sequences share: the sou… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Lightcash: Agentic bookkeeper and VP of finance for startups | — | ↗Not briefed |
hn hn-ai | 29d | Bill LLMbergh a Robot CEO | — | ↗Not briefed |
arxiv arxiv-ai | 29d | AIRRI: Towards a Human-Readable, Machine-Resistant Reading Interface | The rapid advancement of image-to-text extraction systems and large language models (LLMs) poses new challenges for educational assess… | ↗Not briefed |
bluesky bluesky-ai | 29d | 'Topology papers dumped by OpenAI' — the greatest thread in the history of ALGTOP-L, locked by a moderator after 74 posts. | 'Topology papers dumped by OpenAI' — the greatest thread in the history of ALGTOP-L, locked by a moderator after 74 posts. | ↗Not briefed |
bluesky bluesky-ai | 29d | reviewing AI output takes as long as writing it when you factor in the back and forth with the model, the 90% tracks with that | reviewing AI output takes as long as writing it when you factor in the back and forth with the model, the 90% tracks with that | ↗Not briefed |
hn hn-ai | 29d | Lowcart, delivered beauty prices checked at each store's checkout | — | ↗Not briefed |
rss theverge-ai | 29d | Facebook will let you customize your feed algorithm using AI | Facebook's new "Your Algorithm" feature will let you directly add and manage topics you want to see in your feed and Facebook Reels. M… | ↗Not briefed |
bluesky bluesky-ai | 29d | Anthropic said the false murder tip was one of several incidents involving “websites run by US government agencies at the federal, state, a… | Anthropic said the false murder tip was one of several incidents involving “websites run by US government agencies at the federal, sta… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Let Muse book your services and leave you a message | — | ↗Not briefed |
hn hn-ai | 29d | LFM2.5-24B-A2B: Built for Fast On-Device Agents | — | ↗Not briefed |
rss theverge-ai | 29d | Google makes it easier to launch your vibe-coded apps | The next time you vibe-code a custom app with Google's AI Studio, you won't have to connect your phone to a computer to install it. On… | ↗Not briefed |
rss techcrunch-ai | 29d | Google Play now supports private, vibe-coded apps built just for you | Google is making it easier to build and use personal Android apps created with AI, without having to make them available to everyone o… | ↗Not briefed |
hn hn-ai | 29d | Developer pleads with sponsors to pause donations in order to make acc changes | — | ↗Not briefed |
hn hn-ai | 29d | Show HN: Muster, a Mac Menu Bar Switcher for Claude Code, Codex and Cursor | — | ↗Not briefed |
arxiv arxiv-ai | 29d | On Collusion-Free Teaching in Planar Graphs | Non-clashing teaching minimizes teaching complexity among models satisfying the Goldman--Mathias collusion-avoidance criterion. While… | ↗Not briefed |
hn hn-ai | 29d | Gemini 4 Argon for Robotics Blog | — | ↗Not briefed |
arxiv arxiv-ai | 29d | Steertuning Policies from a Few Human Videos | Current recipes for robot learning require massive amounts of robot demonstrations to obtain robust and adaptive policies. In this wor… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Treesum – Replace MS Excel/Project with a Code Editor | # Intro<p>The goal of this software is to make it easy to get the plan that's in your head onto the page. Start with the outline… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Mac equivalent of Nvidia ShadowPlay. Recording missed moments on Mac | (TLDR at the end)<p>Hello everyone,<p>I've been a gamer all my life, and ever since I got a Mac, I've noticed that I am unab… | ↗Not briefed |
bluesky bluesky-ai | 29d | A month ago, OpenAI used a Large Language Model to solve one of math’s “million-dollar” problems. Controversy swirled over their proof’s si… | A month ago, OpenAI used a Large Language Model to solve one of math’s “million-dollar” problems. Controversy swirled over their proof… | ↗Not briefed |
hn hn-ai | 29d | Gemini 4 Argon is now available for everyone | — | ↗Not briefed |
arxiv arxiv-ai | 29d | Unlocking Token Rewards via Training-Free Reward Attribution | In this paper, we propose an extremely efficient, training-free method to extract token-level reward signals directly from an existing… | ↗Not briefed |
bluesky bluesky-ai | 29d | Anthropic conjures another threat that doesn't exist yet to stoke fears of AI apocalypse | Anthropic conjures another threat that doesn't exist yet to stoke fears of AI apocalypse | ↗Not briefed |
bluesky bluesky-ai | 29d | i think the biggest flaw in the series of assumptions made by that anthropic smart tv blog post is that the televisons remain powered on th… | i think the biggest flaw in the series of assumptions made by that anthropic smart tv blog post is that the televisons remain powered… | ↗Not briefed |
hn hn-ai | 29d | Show HN: Lobster – The Skin for Agents | Lobster is an initiative to make sure humans get the best out of an agent. The Lobster skin helps to get a thorough understanding and… | ↗Not briefed |
bluesky bluesky-ai | 29d | “We all know that OpenAI is not doing this “for the good of science”. These pdfs are exhaust fumes from their internal benchmarking effort.” | “We all know that OpenAI is not doing this “for the good of science”. These pdfs are exhaust fumes from their internal benchmarking ef… | ↗Not briefed |
bluesky bluesky-ai | 29d | Impressive, bc my understanding is that the Chinese bar exam is *really hard*. What I’ve been told is that a lot of the Chinese LLM studen… | Impressive, bc my understanding is that the Chinese bar exam is *really hard*. What I’ve been told is that a lot of the Chinese LLM s… | ↗Not briefed |
hn hn-ai | 29d | Popcorn: A security workbench of disposable microVMs | — | ↗Not briefed |
bluesky bluesky-ai | 29d | i seriously have no idea how anyone is using this interface without asking an LLM, this is horrid. no wonder you need certifications for th… | i seriously have no idea how anyone is using this interface without asking an LLM, this is horrid. no wonder you need certifications f… | ↗Not briefed |
bluesky bluesky-ai | 29d | I'm sure there's interesting academic work being done, but for more general purposes the question of whether an LLM has consciousness seems… | I'm sure there's interesting academic work being done, but for more general purposes the question of whether an LLM has consciousness… | ↗Not briefed |
arxiv arxiv-ai | 29d | UniW(A)M: A Unified Mixture-of-Experts Framework for World and World-Action Models | World models (WMs) and world-action models (WAMs) focus on future world prediction and executable action generation, respectively. Alt… | ↗Not briefed |
arxiv arxiv-ai | 29d | A Refresh Interval Does Not Fully Specify a Refresh Schedule: Learning-Rate-Dependent Placement in Kronecker-Preconditioned Language Model Training | Kronecker-preconditioned optimizers recompute a cached curvature eigendecomposition only periodically, and implementations and adaptiv… | ↗Not briefed |
arxiv arxiv-ai | 29d | GeoReader: Learning Latent Geometry for Spatial Reasoning in Vision-Language Models | Vision-language models (VLMs) struggle with metric and multi-view relationships that remain implicit in RGB observations. Geometry fou… | ↗Not briefed |
rss theverge-ai | 29d | Actress Mia Goth says it’s ‘incomprehensible’ that Instagram allowed an AI image of her to spread | Mia Goth at The 32nd Annual Actor Awards. | Variety via Getty Images Earlier this month, news broke that actress Mia Goth had filed… | ↗Not briefed |
arxiv arxiv-ai | 29d | Just Respect the Semantics: Image-Token Compression via Pre-LLM Semantic Variable-Resolution Tokenization and In-LLM Relevance Pruning | Image-token compression must reduce computation without hurting the semantics. We introduce Varest, guided by semantic fungibility: to… | ↗Not briefed |
arxiv arxiv-ai | 29d | RadarGSOcc: Doppler-Aware 3D Semantic Occupancy Prediction Using 4D Radar Tensor with Geometric-Semantic Feature Extraction | 3D semantic occupancy prediction is crucial for autonomous driving, but camera-based methods degrade in low-light conditions and LiDAR… | ↗Not briefed |
arxiv arxiv-ai | 29d | Knowing the User Is Not Knowing What We Established: Grounding Awareness in Personal Agents | Personal-agent memory has largely focused on learning what users know, prefer, and want. However, knowing the user does not necessaril… | ↗Not briefed |
arxiv arxiv-ai | 29d | ProAS: A Progress-Aware Advantage Shaping Framework for Multi-Turn Agentic Reinforcement Learning | Reinforcement learning (RL) has emerged as the primary paradigm for optimizing Large Language Model (LLM) agents, yet efficient critic… | ↗Not briefed |