Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
Researchers released FairLMs, a Python library that unifies fairness measurement and mitigation for language models through explicit declarations of model capabilities and input requirements, according to an arXiv paper …
A new arXiv paper (2609.21425v1) argues that zero-shot time-series forecasting should be governed as an evidence-access claim rather than a training-status condition, since a frozen language model prompted with serialize…
A new arXiv paper (2609.20912v1) reports that RydbergGPT, an autoregressive transformer trained on qubit projective measurement data from interacting Rydberg atom arrays, follows a power-law loss-versus-dataset-size rela…
Attention-Aware Routing (AAR), a method that augments a Mixture-of-Experts router with temporal and spectral features from a sliding window of attention weights, improved GSM8K accuracy by 3.37 percentage points over a r…
RBS-Attention, a training-free sparse-prefill method from a new arXiv paper (2609.20971v1), achieves 20.65x standalone prefill-attention speedup, 11.92x vLLM prefill-attention speedup, and 5.97x end-to-end time-to-first-…
Researchers proposed Causal Latent Revision (CaLR), a framework that reformulates reasoning as constrained latent optimization by adopting a causal topology matrix (CTM) from an expert model and using implicit differenti…
A new arXiv paper (2609.21096v1) introduces a single-pass method that uses Forman-Ricci curvature to analyze attention-graph topology and distinguish hallucinated from non-hallucinated LLM responses. The authors report c…
A clinician-grounded evaluation platform called InterviewPlayground, built around a memory-augmented patient simulator for open-ended AI interviewing, was tested in a pilot with 6 clinicians in a 25-minute assessment aga…
A new arXiv paper (2609.21157v1) reports that an LLM-agent workflow called Agent-based HLS with RTL Refinement (AHRR) achieved a 2.6x geometric-mean speedup over Direct RTL Design across an 11-task benchmark suite. The s…
Researchers introduced CogGym, a scalable framework grounded in cognitive science that standardizes 258 cognitive experiments from 100 papers into a task-agnostic Experiment Markup Language (EML) to compare human and mac…
A study of a production analytics LLM agent serving tens of thousands of monthly active users found that multidimensional 2PL adaptive testing delivered the best score fidelity, executing 200 questions — 38.5% of a full …
A multimodal reasoning framework using Task-Aware Group Relative Policy Optimization (GRPO) achieved an Overall Score of 83.24 on the official PCBA Standard-to-Real Grand Challenge leaderboard, according to arXiv paper 2…
Z.ai released GLM-5.3-Flash on August 26, 2026, a 320-billion-parameter open-weight Mixture-of-Experts model that activates only 18 billion parameters per token and supports a one-million-token context under the MIT lice…
An engineer argues that swapping LLM model IDs upon deprecation without task-specific evaluation is a common cause of silent production degradation, since benchmark gains do not guarantee better performance on a team's o…
A developer has laid out a detailed scenario in which a bored 8-year-old, using a school-district AI assistant and his mother's delivery account, orders parts for a convergent implosion device that creates a slow-moving …
Five frontier-class AI models shipped between September 1 and September 10, 2026 — Anthropic's Claude Fable 5.1, Google's Gemini 3.8 Flash, Meta's Muse Spark 1.3, OpenAI's GPT-6 Astra, and DeepSeek's V4.1-Flash — creatin…
Modal, Fireworks AI, and Baseten are offering hosted inference for Moonshot AI's Kimi K3 at roughly one-tenth the cost of direct access, using Nvidia GB300 NVL72 systems and AMD MI350X and MI355X accelerators that US exp…
Large language models are increasingly used as automated reviewers and as assistants to human reviewers in scientific evaluation, according to a study on recursive AI peer review. As model-generated reviews enter public …
The open-source Python package ora-grounding, released under the MIT license and installable via `pip install ora-grounding`, adds deterministic post-response grounding checks plus cross-family adversarial review for LLM…
A Hugging Face forum user identified as mhutchw asked whether provider-managed KV caching is sufficient for agent swarms and long-running agents, citing complaints from inference engineers and startups about the lack of …