AI Research news and analysis on Web Pulse: 22651 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
An INT8 quantization experiment on a TinyML ECG arrhythmia detection model cut its size from 87.5 KB in FP32 to 34.6 KB, a roughly 60% reduction, while reported accuracy held at 93.8%, according to the Beat2Bit project w…
A Thai AI practitioner's comparative analysis finds that locally trained Thai language models outperform global frontier models on Thai-specific tasks such as document and handwriting recognition, legal citation, Isan di…
Google DeepMind is reportedly near achieving Recursive Self-Improvement (RSI), according to an unverified report, with Google co-founder Sergey Brin pushing resource allocation toward RSI in Google's AI efforts around Au…
A survey of Thai-language AI resources on Hugging Face highlights underused datasets and models, including the OpenThaiGPT evaluation dataset and the WangchanThaiInstruct multi-turn conversation set, alongside SCB 10X's …
A developer behind the Edge0 project demonstrated a 35B-parameter language model running on an iPhone with only 1–2.5 GB of peak memory by using Mixture of Experts (MoE) routing to load only the needed expert weights fro…
Google Research released TimesFM-3, a 330-million-parameter forecasting model that analyzes time series data alongside related information and known future events such as sales promotions or weather forecasts. Rather tha…
An analysis of 623 million real-world code changes from 2023 to 2026 by GitClear and GitKraken found code duplication rose 81% while code reuse fell 70% and error masking climbed 47%, producing code that is locally corre…
A developer at cognee outlined a structured test for evaluating coding-agent memory systems, using two fictional projects, Atlas and Beacon, to check whether an agent retains a correction, preserves historical commands, …
OpenAI's GPT-6 Astra solved the final remaining problem in the elite FrontierMath Tier 4 benchmark, taking the model from a sub-2% success rate to total saturation in under two years. In response, twenty-five Fields Meda…
OpenAI's claim to have solved the Navier-Stokes Millennium Prize Problem in 88 hours using roughly 10,000 AI agents and 130 billion output tokens is being challenged on the grounds that the work may have leaned on prior …
OpenAI's latest AI model solved a Millennium Prize Problem, a puzzle carrying a $1m reward that had defied human mathematicians for decades, according to The Guardian. The company deployed 10,000 autonomous AI agents on …
Lance Eliot, a Forbes contributor, reported on using AI personas to perform the Creativity Triangle Task as a way to explore Langerian Mindfulness, invoking the Langerian Mindfulness Scale (LMS). The AI Insider analysis …
An independent developer has open-sourced Beatriz Epistemic Gate, a lightweight defensive proxy that detects data poisoning during LLM fine-tuning by verifying generated text against an immutable anchor corpus via embedd…
Three Thai open-weight LLM projects — SCB 10X's Typhoon, the Thai AI Entrepreneur Association's OpenThai, and VISTEC's Pathumma — offer locally runnable, commercially usable alternatives to foreign models, each different…
A developer built a 5-layer, 512-wide TinyStories language model that runs natively on Apple's M4 CPU by treating ARM Neon instructions—including XOR, popcount, and SDOT—as trainable architectural primitives rather than …
A2ABreak, the first systematic security analysis of the Linux Foundation-governed Agent2Agent (A2A) protocol, uncovered 11 new vulnerabilities that a specification-compliant adversary can exploit without any implementati…
A historical survey traces the first physical and computational reinforcement learning machines built between 1948 and 1954, beginning with Norbert Wiener's 1948 Cybernetics and its negative-feedback principle, which the…
DeepSeek released DeepSeek v4.1-Flash, a 763B-parameter model with a novel causal encoder-decoder architecture that splits 8B parameters to prefill and 16B to decode, per the model's tech report on Hugging Face. The rele…
At JacHacks SF, developers Tirth Suba and Aarnav Gutti built Trails, a multi-agent research system inspired by ant foraging behavior in which five AI agents coordinate through a shared evidence graph rather than direct m…
A developer analysis of Artificial Analysis benchmarking data finds that GLM-5.3-Flash scores 42 at $0.25 per task while Kimi K3 scores 44 at $2.00 per task, an eight-fold cost gap for a two-point capability difference. …