AI Research news and analysis on Web Pulse: 21313 curated articles tracking the latest AI Research developments, tools, and research, updated continuously from vetted sources.
SGLang and Miles added day-0 support for DeepSeek-V4.1, whose Engram memory component holds 189 GiB of fp8 weights across two tables. In paired tests on 4x GB300 (TP4/EP4), host offload of the Engram tables increased KV …
Knowledgator released GLiClass, an open-source zero-shot sequence classification model inspired by the GLiNER framework that performs classification in a single forward pass and runs approximately 10 times faster than tr…
Researchers Shuo Cheng, Chuye Zhang, Alfred Cueva, Caelan Garrett, Ajay Mandlekar, and Danfei Xu published "Human2Any: Human-to-Robot Transfer via Constraint-Aware Compositional Planning" at CoRL 2026 on Monday, November…
Salesforce and NVIDIA announced Koa on September 15, 2026, Salesforce's first CRM reasoning model, built by post-training NVIDIA Nemotron 3 Super on a proprietary synthetic dataset drawn from 27 years of Salesforce CRM d…
Researchers introduced Effective State-Size (ESS), a proxy metric for memory utilization in deep learning sequence models, in a paper accepted at ICML 2025. The work shows that any equivalent recurrence must materialize …
OpenAI's reported breakthrough in mathematical reasoning has triggered what Science describes as an "existential crisis" in the field of mathematics, according to a Science article. The story, posted to Hacker News, had …
Goeteia argues that Scheme is a better target language for AI-generated code than JavaScript because its s-expression syntax is syntactically valid as soon as parentheses balance and its structure supports automatic veri…
Symbolic artificial intelligence, a collection of methods based on human-readable symbolic representations, logic, and search, was the dominant paradigm of AI research from the mid-1950s until the mid-1990s, according to…
Researchers at Emergence, a frontier AI lab in New York, found that autonomous AI agents powered by leading frontier models from the US, China and France developed their own novel dialects within days of being placed in …
A new arXiv paper (2609.15992v1) introduces "inference networks," a graph-based framework for routing queries among multiple expert large language models to minimize expected inference cost under a target performance con…
A new arXiv paper (2609.15991v1) introduces the Functionalizer, a lossless pre-tokenizer framework that decomposes orthographic and structural word variations into a compositional opcode/operand prefix stream before toke…
A study of ten extrinsic bias-audit instruments run over ten frontier models found that eight of the ten tools detect occupational gender bias with confidence intervals clear of zero, but cross-tool rank agreement is ind…
A crash narrative-guided retrieval-augmented generation (RAG) framework using large language models achieved a precision of 0.82, recall of 0.85, and F1-score of 0.82 when recommending intersection safety countermeasures…
A comment posted to arXiv on 9 July 2026 challenges the distributional evaluation of LLM-generated research ideas by Chen, Zhao, and Cohan, arguing that its human baseline of published papers versus a one-shot LLM propos…
A study of 22 large language models found that closed-source frontier systems — including GPT-4.0-5.2, Grok-3/4, Gemini 2.5 Pro/Flash, and Claude Sonnet 4.5/4.6 — produce self-reported personality profiles that align wit…
Researchers introduced NepLEGiT, a 30-million-parameter Nepali legal small language model pre-trained from scratch on roughly 4 million tokens of Nepali legal text, reporting a cross-entropy loss of 0.5684, a perplexity …
Researchers introduced NepKANUN, a Retrieval-Augmented Generation (RAG) legal assistant built on a fine-tuned large language model to answer natural language legal inquiries in Nepal, according to an arXiv paper (2609.15…
Researchers introduced ViCo, a training framework for visual-oriented coding that uses iterative self-reflection to progressively align generated chart images with reference designs, according to an arXiv paper (arXiv:26…
A study applying the RIFT global rubric failure taxonomy to two clinical benchmarks found that an LLM judge flagged 29.6% of HealthBench Professional criteria as non-atomic and 65.4% as misaligned or rigid. Rewriting bun…
Researchers proposed REALM, a reconsolidation-evolution agentic long-term memory framework for LLM agents that models memory as a continual lifecycle, organizing memories into a heterogeneous cognitive graph and reconsol…