Daily AI News — 2026-08-25

Generated at 2026-08-25T06:11:01.059413-07:00 by a GitHub Actions GitOps pipeline.

Top stories

• LLM4LLM: Bridging Kernel Benchmarks and Real Deployment via Closed-Loop Agentic Optimization — arXiv cs.AI, 2026-08-25 04:00 UTC. Score 10.1 — arXiv:2608.21836v1 Announce Type: new Abstract: Large language models have become increasingly capable agents for low-level code and kernel optimization, but isolated kernel benchmarks provide only a proxy for the deployment behavior that matters in language-model inference. We identify a benchmark-to-deployment gap: candidate kernels that appear correct and fast in standalone harnesses can exhibit different performance, safety, or phase behavior after integration into a real inference workload. We introduce LLM4LLM, a deployment-aware closed-loop optimization framework that starts from a target inference script, extracts phase-aware optimiz…
• Can LLMs Truly Forget? Revealing Unlearning Gaps Through Adversarial Evaluation — arXiv cs.CL, 2026-08-25 04:00 UTC. Score 10.1 — arXiv:2608.21606v1 Announce Type: new Abstract: Machine unlearning aims to remove the influence of targeted training data from a model while preserving its remaining capabilities, but evaluating whether such information has truly become inaccessible remains challenging. Existing benchmarks primarily assess unlearning under clean, non-adversarial queries, leaving open whether information that appears forgotten can still be recovered through strategic prompting. We address this gap through a unified evaluation of prompt-based and fine-tuning-based unlearning methods on TOFU using Llama-3.2-3B-Instruct, followed by an adversarial robustness eva…
• Spyre-Accelerated Retrieval-Augmented Generation on IBM LinuxONE: A Cloud-Native Architecture for Secure, High-Throughput Enterprise AI Inference — arXiv cs.AI, 2026-08-25 04:00 UTC. Score 10.0 — arXiv:2608.21393v1 Announce Type: new Abstract: Running large language models inside enterprise environments has always bumped up against a practical wall: the data lives in one place, the AI horsepower sits somewhere else, and moving sensitive records between the two creates real headaches around latency, security, and regulatory exposure. IBM's Spyre accelerator PCIe inference card built for LinuxONE and the broader IBM Z family changes that equation. In this paper we lay out a six-subsystem RAG architecture that runs entirely on IBM LinuxONE, using Spyre for generative inference, the Telum II on-chip accelerator for lightweight classifica…
• Beyond Success and Failure: Length-Aware Contrastive Learning for GUI Agents — arXiv cs.AI, 2026-08-25 04:00 UTC. Score 9.2 — arXiv:2608.21830v1 Announce Type: new Abstract: Graphical User Interface (GUI) agents powered by Multimodal Large Language Models (MLLMs) have shown strong potential for automating tasks across diverse digital environments, where reinforcement learning (RL) has become a dominant training paradigm. However, widely used methods such as Group Relative Policy Optimization (GRPO) suffer from reward-gradient misalignment, leading to inefficient and unstable optimization. Recent work addresses this issue by reformulating RL with verifiable rewards (RLVR) as contrastive or classification-based objectives, which improve stability by eliminating probl…
• Traceable Spectral Inference via Influence Functions: Efficient Data Attribution and Error Proxies for the Ariel Mission — arXiv stat.ML, 2026-08-25 04:00 UTC. Score 9.2 — arXiv:2608.23458v1 Announce Type: cross Abstract: Interpretability is critical for machine learning models deployed in scientific space missions such as ESA's Ariel, where ground truth is unavailable during operations and physical plausibility must be assessed. While most explainable AI methods focus on feature attribution, this work investigates training data attribution through influence functions and introduces three key contributions for operational spectroscopy pipelines. First, influence is reformulated in terms of prediction rather than loss, enabling label-free deployment. Second, by leveraging the closed-form ridge solution of an Ex…
• CausalSmith: A Formally Grounded, Self-Improving Agentic Framework for Automated Research in Causal Inference — arXiv stat.ML, 2026-08-25 04:00 UTC. Score 9.2 — arXiv:2607.22511v3 Announce Type: replace Abstract: Automating theoretical research is constrained not only by the generation of candidate results, but also by their reliable evaluation. A common approach is to close the research loop with a large language model (LLM) reviewer. However, such reviewers remain empirically unreliable: they may accept fabricated papers and detect them at rates close to chance (Bad Scientist, 2025). We present CausalSmith, a framework for automated theoretical research in causal inference grounded in the Lean proof assistant. CausalSmith combines Causalean, a foundational Lean library for causal inference contain…
• K-Bench: measuring model performance on real scientific agent requests — arXiv cs.AI, 2026-08-25 04:00 UTC. Score 8.4 — arXiv:2608.21601v1 Announce Type: new Abstract: Benchmarks for scientific artificial intelligence are mostly written to be scored: multiple-choice questions, curated agent tasks with reference solutions, or simulators with a known generative structure. Real scientific requests arrive differently. They are underspecified, they carry attachments, and lack ground truth. We report K-Bench 01, an evaluation built from first-turn requests sampled from live user traffic on K-Dense Web and run end to end by nine frontier models in identical sandboxes, yielding 1,602 completed agent runs. Three blinded language-model judges scored every run against a…
• Training a Knowledge Base: Supervised Structure Learning for Agent-Curated Document Stores — arXiv cs.CL, 2026-08-25 04:00 UTC. Score 8.4 — arXiv:2608.21829v1 Announce Type: new Abstract: Retrieval-augmented generation treats the document store as a frozen input, and the systems that instead let an agent curate one never measure what curation does to the store. We invert the framing: the knowledge base is the model. A training agent answers a supervised question against the current store, is shown the gold, then edits the store; an unchanged reader is later examined on a frozen snapshot under a fixed action budget. Where offline graph construction is unsupervised, (question, answer) pairs are our labels -- and that supervision is what makes the structure cheap. Per point of corp…
• GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding — arXiv cs.CL, 2026-08-25 04:00 UTC. Score 8.4 — arXiv:2608.21832v1 Announce Type: new Abstract: Computer-use agents ground natural-language instructions in screenshots to locate interface elements, yet existing benchmarks do not isolate whether models bind relational language to the correct element. We introduce GUI-Primitives, a 994-item benchmark of contrastive instruction pairs over seven spatial relations in graphical user interfaces (left/right, above/below, containment, alignment, proximity, list ordinal, occlusion). Each pair holds the screenshot and anchor fixed while changing the relation expression, so the correct target moves between two designated candidates. Five annotators v…
• Improving Few-Step Language Flows with Untied Self-Conditioning — arXiv cs.CL, 2026-08-25 04:00 UTC. Score 8.4 — arXiv:2608.22244v1 Announce Type: new Abstract: Flow-matching language models refine all token positions in parallel and can trade sampling steps for latency, yet generation quality still degrades sharply with few sampling steps. We trace a source of this degradation to a train--inference mismatch in previous-prediction self-conditioning: during training, the self-conditioning input is computed from the current noisy state with no intervening solver step; during sampling, the solver folds the previous prediction into the latent before that same prediction reappears as the explicit self-conditioning input. This coupling, absent during trainin…

Signals to watch

• Most represented sources: arXiv cs.AI (4), arXiv cs.CL (4), arXiv stat.ML (4), TechCrunch AI (3).
• Recurring themes: model (13), agent (11), inference (7), training (7), benchmark (5), openai (3), multimodal (2), llama (1).
• Review the raw JSON output in data/items/ if you want to audit scoring or feed coverage.

All links

• LLM4LLM: Bridging Kernel Benchmarks and Real Deployment via Closed-Loop Agentic Optimization — arXiv cs.AI
• Can LLMs Truly Forget? Revealing Unlearning Gaps Through Adversarial Evaluation — arXiv cs.CL
• Spyre-Accelerated Retrieval-Augmented Generation on IBM LinuxONE: A Cloud-Native Architecture for Secure, High-Throughput Enterprise AI Inference — arXiv cs.AI
• Beyond Success and Failure: Length-Aware Contrastive Learning for GUI Agents — arXiv cs.AI
• Traceable Spectral Inference via Influence Functions: Efficient Data Attribution and Error Proxies for the Ariel Mission — arXiv stat.ML
• CausalSmith: A Formally Grounded, Self-Improving Agentic Framework for Automated Research in Causal Inference — arXiv stat.ML
• K-Bench: measuring model performance on real scientific agent requests — arXiv cs.AI
• Training a Knowledge Base: Supervised Structure Learning for Agent-Curated Document Stores — arXiv cs.CL
• GUI-Primitives: Diagnosing Spatial Reasoning Failures in Vision-Language GUI Grounding — arXiv cs.CL
• Improving Few-Step Language Flows with Untied Self-Conditioning — arXiv cs.CL
• Guidance for Prior Change via Density Ratio Estimation — arXiv stat.ML
• Valor, Point72 back General Intuition at $6B valuation as AI startup pushes into robotics — TechCrunch AI
• Gauss--Hermite Quadrature for Gaussian-Mixture Entropy with an Action-Space Hermite Surrogate — arXiv stat.ML
• ‘The world seems to be ready’: An interview with OpenAI head of product Thibault Sottiaux — TechCrunch AI
• OpenAI subpoenaed by Alabama AG over Hugging Face hack — The Verge AI
• Introducing new Ray capabilities on SageMaker HyperPod — AWS Machine Learning Blog
• Building a restaurant telephony AI host with Amazon Connect — AWS Machine Learning Blog
• OpenAI is building AI agents for everything. Will everyone use them? — TechCrunch AI

Pipeline metadata

• Selected items: 18
• AI summarization: deterministic fallback
• Source data: data/items/ in this repository