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 |
|---|---|---|---|---|
rss techmeme | 19d | Town, which develops enterprise AI assistants, raised a $108M Series B at a $1B valuation, after announcing a $55M Series A in June (Dominic-Madori Davis/TechCrunch) | Dominic-Madori Davis / TechCrunch: Town, which develops enterprise AI assistants, raised a $108M Series B at a $1B valuation, after an… | ↗Not briefed |
bluesky bluesky-ai | 19d | *checks in on LinkedIn* *40% posts on ArtCraft* *30% posts on fake flying cars* *30% posts on fake Anthropic/SpaceX stock "philanthropy"* *… | *checks in on LinkedIn* *40% posts on ArtCraft* *30% posts on fake flying cars* *30% posts on fake Anthropic/SpaceX stock "philanthrop… | ↗Not briefed |
arxiv arxiv-ai | 19d | EDMA: Entropy-Driven Multimodal Answering | Multimodal question answering (MMQA) is based on integrating heterogeneous data sources, selectively leveraging relevant modalities, a… | ↗Not briefed |
bluesky bluesky-ai | 19d | el llm que genera la temática de las pelis en netflix se pasó de tópico | el llm que genera la temática de las pelis en netflix se pasó de tópico | ↗Not briefed |
arxiv arxiv-ai | 19d | PORA: Learning When More Reasoning Pays Off for LLM Agents | LLM agents perform tasks over multiple steps and revise their actions based on feedback. As the total reasoning budget is often limite… | ↗Not briefed |
bluesky bluesky-ai | 19d | Before he was arrested, the Saudi national accused of bugging a restricted OpenAI conference room, had allegedly been in touch with a membe… | Before he was arrested, the Saudi national accused of bugging a restricted OpenAI conference room, had allegedly been in touch with a… | ↗Not briefed |
arxiv arxiv-ai | 19d | PROVISO: Provenance-Informed Action Selection for Secure LLM Agents | LLM agents act on the external world through tools and third-party skills, so a single executed action can cause irreversible harm. Ex… | ↗Not briefed |
bluesky bluesky-ai | 19d | Anything generated by an LLM will still be bound to the same coding fallacies and looping based on the ingestion data without the context/u… | Anything generated by an LLM will still be bound to the same coding fallacies and looping based on the ingestion data without the cont… | ↗Not briefed |
arxiv arxiv-ai | 19d | Joint Constraint Aggregation for Qualifier-Sensitive Retrieval-Augmented QA | Retrieval-augmented QA can fail even when the correct answer is in the retrieved evidence. This can happen when a competing answer sat… | ↗Not briefed |
hn hn-ai | 19d | RuneScape Client API | — | ↗Not briefed |
arxiv arxiv-ai | 19d | Beyond Two Tracks: Speaker-Anchored Full-Duplex Dialogue Recovery | Spoken interaction is moving from alternating turns to full-duplex conversation, where an agent listens while speaking. Learning this… | ↗Not briefed |
bluesky bluesky-ai | 19d | I put a new website (domain) on the internet, and in less than an hour it started getting fully crawled by openai gptbot | I put a new website (domain) on the internet, and in less than an hour it started getting fully crawled by openai gptbot | ↗Not briefed |
bluesky bluesky-ai | 19d | While reading this, it’s worth noting that Microsoft has made using these models almost inescapable in their operating system, most popular… | While reading this, it’s worth noting that Microsoft has made using these models almost inescapable in their operating system, most po… | ↗Not briefed |
arxiv arxiv-ai | 19d | RGH-GNN : Risk-Gated Heterogeneous Graph Neural Network for Financial Fraud Detection | Credit card fraud is a significant problem for financial institutions, resulting in substantial monetary loss each year. Fraud pattern… | ↗Not briefed |
arxiv arxiv-ai | 19d | This is SParTA: A modular, efficient and scalable transformer architecture for jet-tagging | We present the Sparse Particle-interaction Transformer Architecture (SParTA), a modular, scalable, and efficient transformer for jet t… | ↗Not briefed |
bluesky bluesky-ai | 19d | Developers and vibe coders are using AI tools like Claude Code to get rid of sponsored ads on their smart TVs, disable tracking features, a… | Developers and vibe coders are using AI tools like Claude Code to get rid of sponsored ads on their smart TVs, disable tracking featur… | ↗Not briefed |
rss techmeme | 19d | Developers and vibe coders are using AI tools like Claude Code to get rid of sponsored ads on their smart TVs, disable tracking features, and more (Sarah Perez/TechCrunch) | Sarah Perez / TechCrunch: Developers and vibe coders are using AI tools like Claude Code to get rid of sponsored ads on their smart TV… | ↗Not briefed |
bluesky bluesky-ai | 19d | Using an LLM like this without any actual oversight and coding knowledge is dangerous as fuck. You might open doors to even more uninvited … | Using an LLM like this without any actual oversight and coding knowledge is dangerous as fuck. You might open doors to even more uninv… | ↗Not briefed |
arxiv arxiv-ai | 19d | Omni-Woodpecker: Hallucination Correction via Evidence-Prior Rebalancing for Omni LLMs🔧 Modal | Recent Omni LLMs (OLLMs) have demonstrated remarkable capabilities in joint audio-visual reasoning, yet cross-modal hallucinations rem… | ↗Not briefed |
hn hn-ai | 19d | Show HN: Quorus – Claude Code and Codex in one room, reviewing each other's work | — | ↗Not briefed |
arxiv arxiv-ai | 19d | LOB-Impact Benchmark: Emergent Market Impact in Generative Limit Order Book Models | Realistic market impact is essential for generative limit order book (LOB) models to serve as world models for trading agents. If a si… | ↗Not briefed |
bluesky bluesky-ai | 19d | By now it's clear most LLM service providers' plan to become profitable is the Walmart gambit – i.e., secure a monopoly by offering your se… | By now it's clear most LLM service providers' plan to become profitable is the Walmart gambit – i.e., secure a monopoly by offering yo… | ↗Not briefed |
arxiv arxiv-ai | 19d | A general-purpose System One decision model for clinical records | Clinical care, health-system operations and biomedical discovery depend on extracting reliable answers from patient records, yet apply… | ↗Not briefed |
arxiv arxiv-ai | 19d | SciFigBench: A Benchmark for Scientific Method Figure Understanding | Scientific method figures often provide researchers with a clear visual entry point into a paper's methodology and core contributions,… | ↗Not briefed |
arxiv arxiv-ai | 19d | LEAF: LUT-aware Efficient Attention on FPGAs | We introduce Lookup-table (LUT)-aware Efficient Attention on FPGAs (LEAF). LEAF is the first nanosecond-scale LUT-aware transformer fo… | ↗Not briefed |
arxiv arxiv-ai | 19d | ClockMark: Adaptive and Host-Dependent Watermarking for Large Language Models | Large language model (LLM) watermarking operates on the next-token distribution that evolves during generation. In streaming generatio… | ↗Not briefed |
arxiv arxiv-ai | 19d | Graph Neural Networks for Astrophysics and Cosmology | Physics seldom makes breakthroughs when we consider objects only in isolation. Instead we model and measure how objects sit among thei… | ↗Not briefed |
arxiv arxiv-ai | 19d | Text-to-Image Generative Modeling Without Language Pretraining | Modern text-to-image (T2I) systems use internet-scale language pretraining for text-conditioned image generation. We ask whether langu… | ↗Not briefed |
arxiv arxiv-ai | 19d | OpenVLHarness: Real World Visual Agentic Tool-Use with Multimodal Memory | Multimodal large language models (MLLMs) have made rapid progress on visual reasoning, yet many tasks require evidence that is difficu… | ↗Not briefed |
arxiv arxiv-ai | 19d | Can It Actually Work? Synthesizing Functional CAD Mechanisms | While LLM agents can generate visually plausible CAD assemblies, these designs often fail to fulfill their intended physical functions… | ↗Not briefed |
arxiv arxiv-ai | 19d | Reinforcement Learning with Visually Attributable On-Policy Self-Distillation | Multimodal reinforcement learning with verifiable rewards (RLVR) provides reliable task-level supervision, yet offers limited guidance… | ↗Not briefed |
arxiv arxiv-ai | 19d | CoopUQ: Robust Multi-Robot Social Navigation via Uncertainty-Aware Cooperative Forecasting | Reinforcement learning enables mobile robots to make fast and scalable navigation decisions in constrained human-shared environments.… | ↗Not briefed |
arxiv arxiv-ai | 19d | Where is Physical RSI capability from? An Observational Scaling Law of LLM-as-policy | In this paper, we systematically investigate the mechanisms, scaling behavior, and generalization of physical Recursive self-improveme… | ↗Not briefed |
arxiv arxiv-ai | 19d | WeAct: Dynamic Weight Switching as Agent Actions | Long-horizon agents must learn heterogeneous capabilities from interleaved trajectories. Updating shared parameters to learn new capab… | ↗Not briefed |
arxiv arxiv-ai | 19d | Train An Imaginative Agent | A seemingly reasonable push in Sokoban can make the puzzle unsolvable. To avoid such mistakes, an agent has to anticipate the conseque… | ↗Not briefed |
arxiv arxiv-ai | 19d | AnyGoal: On World Model Generalization Beyond Expert Rollouts | The goal of a world model is to capture the dynamics of an environment. This has the great appeal that it allows agents to plan and si… | ↗Not briefed |
hn hn-ai | 19d | Show HN: Open-source early-reading reader with phonics help🔧 ChatGPT | My kid is 6 y/o and she is learning to read. She's moved through the basic phonics with Teach Your Monster to Read but was s… | ↗Not briefed |
arxiv arxiv-ai | 19d | LatComm: Reusable Context Compression for Latent Communication between Large Language Models | Systems that combine several large language models (LLMs) often use one context for many questions. When the context is passed between… | ↗Not briefed |
arxiv arxiv-ai | 19d | Newton-Based Methods for Bilevel Optimization Beyond Lower-Level Strong Convexity | Bilevel optimization provides a framework for machine learning tasks such as hyperparameter optimization, data reweighting, and meta-l… | ↗Not briefed |
hn hn-ai | 19d | Pentoniq Tube Synth for macOS | — | ↗Not briefed |
arxiv arxiv-ai | 19d | Discrete Neural Interlingua for Ontology-Agnostic EHR Modeling | EHR foundation models promise scalable clinical prediction across health systems, yet their deployment is constrained by a tokenizatio… | ↗Not briefed |
bluesky bluesky-ai | 19d | I couldn’t agree more- give it to NASA and crest a similar model for AI. | I couldn’t agree more- give it to NASA and crest a similar model for AI. | ↗Not briefed |
arxiv arxiv-ai | 19d | Same Simulation, Different Evidence: Where Retrieved Tool Output Costs LLM Agents Accuracy🔧 GPT-5 | LLM agents that compute the right answer can still answer wrongly. Agent harnesses increasingly keep long tool outputs out of the cont… | ↗Not briefed |
arxiv arxiv-ai | 19d | Kronecker Delta Rule for Long Context Linear Attention | We propose KronLSTM, a linear attention architecture based on the Kronecker Delta Rule, a novel update rule for associative matrix mem… | ↗Not briefed |
arxiv arxiv-ai | 19d | Supervised Learning with Quantum Hamiltonian Embeddings | Machine learning is widely regarded as a promising application in which quantum computing could offer a computational advantage. Howev… | ↗Not briefed |
arxiv arxiv-ai | 19d | Competing Learning in Hybrid Architectures | Many modern language models have adopted hybrid architectures that incorporate both linear-time recurrence and softmax attention. Mode… | ↗Not briefed |
arxiv arxiv-ai | 19d | ProteinGeoBench: Frontier Structural-Biology Agents Are Fluent at Lookup but Lost in Geometry | Frontier agents retrieve structural metadata well and compute geometry from coordinates less reliably. We ask how much of a benchmark… | ↗Not briefed |
arxiv arxiv-ai | 19d | Agent World Model in the Wild: Computer-Use Agent Environment Synthesis | Agent learning is shifting from learning from human data toward learning through interaction with the world. However, scaling such lea… | ↗Not briefed |
arxiv arxiv-ai | 19d | KEEP-Bench: Evaluating Evidence Preservation in Context Compression | Context compression improves long-context efficiency but may introduce faithfulness issues through an information bottleneck. Existing… | ↗Not briefed |
arxiv arxiv-ai | 19d | RefCraft: Benchmarking Visual-Reference-Conditioned Agentic Generation through Web Development | Beyond basic tool use and coding skills, effectively incorporating visual references into agentic generation is equally essential for… | ↗Not briefed |