AI news this week: 56 source-linked stories
A dated list of 56 AI stories with links to their original sources.
Stories published this week, linked to their original sources. The underlying claims have not been independently verified.
Format-Aware Fusion for Fast FP4 Pretraining
Four-bit floating-point (FP4) Tensor Cores accelerate matrix multiplication, but scale computation, operand packing, layout construction, and saved backward state can erase the gain.
Constant-Memory Recall: Learned Associations in a Fixed Matrix State
Fixed-size recurrent memory limits storage growth during inference, but successful recall depends on the task and training.
How Far is Adam from Natural Gradient Descent?
Adam’s geometric
"very likely" Means "uncertain"? How LLMs Diverge from Humans in Linguistic Uncertainty Quantification
Humans express uncertainty verbally via markers (e.g., "possible," "likely"), yet most LLM uncertainty quantification (UQ) relies on costing likelihood- or consistency-based signals.
Energy Time-Series Imputation with Differentially Private Diffusion Models via Clipping-Aware Objective Conditioning
Diffusion models trained with
The Hidden Costs of 99% Accuracy: A Trustworthiness Audit of the Telco Customer Churn Benchmark
Customer churn prediction on the IBM Telco Customer Churn benchmark (n = 7,043) routinely reports test accuracies above 95%, with the most cited published study reporting 99.01%.
Fast Polynomial Transcendentals for LLMs
Graphics processing unit (GPU) generations scale matrix, special-function, and memory pipelines at different rates, so kernel bottlenecks move as hardware evolves.
SW-KAN: Kolmogorov-Arnold Networks with Stieltjes-Wigert q-Orthogonal Polynomials
Kolmogorov-Arnold Networks
The Null Is the Hard Part: Exact Tests for Memorization in Generative Models
Memorization
Uncertainty-Aware Learning from Multi-Expert Interval Targets
Many machine learning (ML) applications rely on expert labels, and qualified experts may provide different but plausible interpretations of the same observation.
Heavy-Tailed Memory Traces in Long-Horizon Language Agents
Long-horizon language agents increasingly rely on external memory as a frozen world model, yet current memory systems are usually judged only by task success or token cost.
What Do Rationales Communicate? A Message-Intervention Study in Role-Specialized QA
On 40
EviGraph: Proof-Carrying Selective Recommendation over Temporal Public-Service Knowledge Graphs
Public-service recommendations require evidence that matches the requested service, scope, and date.
K-Dense BYOK: An Open-Source AI Research Assistant That Runs Locally and Keeps a Hash-Chained Lab Notebook
K-Dense BYOK
Measuring the Microtask Eligibility Gap: When Is an Off-the-Shelf SLM Enough for an Agent Harness?
Agent harnesses increasingly want to run small language models (SLMs) on the microtasks around a frontier large language model (LLM) planner: auto-approving shell commands, writing memory, selecting tools, ranking past turns.
When Do Causal World Models Help Modular LLM Agents
LLM agents
Build2SPARQL: A Large-Scale Text-to-SPARQL Benchmark Dataset for Building Knowledge Graph Querying
Build2SPARQL
Characterizing a Configuration Where Inference-Time PRM-Pruned Fragment Grafting Is Inert: Evidence from Three Reasoning LMs
PRM-
NVIDIA DGX Spark 64GB Gives Developers More Ways to Build and Scale Local AI
NVIDIA DGX
Scientific Agents: Evaluating Profession-Specific System Prompts on Scientific Tasks
Detailed profession-
MoFlow: Multi-Objective Agentic Workflow Generation
MoFlow
Productive, Durable, Fungible: How NVIDIA AI Factories Maximize Return on Investment
AI factories are
AI Agents are Vulnerable to Radicalization
Large language models (LLMs) can influence people’s beliefs, yet little is known about whether and how they can manipulate each other.
CARAT: Do Materials LLMs Reason or Recite?
When a materials LLM answers a question about crystal structure, does it reason from the structure or copy an answer already printed in its input?
How NVIDIA GPUs Help Accelerate OpenAI’s GPT-6 Astra Ultrafast
GPT-
Improving OCR Faithfulness via Gated and Attenuated On-Policy Distillation
Vision-language models may rewrite anomalous text in images into linguistically plausible expressions, compromising OCR transcription faithfulness.
AREX-2: Advancing Self-Improving Agents through Long-Horizon Reflective Tasks
We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test time.
Examining Variation in How Guided AI Tutors Resolve Student Impasses
When a student is stuck, a tutor faces the assistance dilemma: help given too early can hinder productive struggle, while help withheld too long leaves the student in a frustrating, persistent impasse (i.e., wheel spinning).
Self-discovering RL in the Era of Experience: Is Learning History an Asset or a Burden?
The pursuit of recursive self
PowerZooJax: A JAX-based Power System Benchmark for Reinforcement Learning
Power system operation is a safety-critical sequential decision-making problem, making it a natural testbed for reinforcement learning (RL).
Representational Simplicity and Circuit Size Dissociate in a Threshold-Dependent Way: A Controlled Test via Adversarial Training
Sparse-autoencoder decomposability and concentrated feature attribution are increasingly treated as evidence that a model’s computation is easier to reverse-engineer.
SAGE: A Statistical Acceptance Gate for Self-Evolving Agents
Large Language Model (LLM)-based agents increasingly self-evolve by editing a persistent skill document that encodes their workflow, tool-use rules, and decision logic.
More Programs or More Rolls? Separating Coverage from Specialization in LLM Harnesses
Automated generation of LLM harnesses promises to improve inference through task specialization.
Is Human-Readable Text Necessary for Effective LLM Fine-Tuning?
Is human readability necessary for effective fine-tuning of large language models?
Mirror-Score: Calibrated, Inference-only Scoring Exposes the Limits of Sequence-compatibility Ranking in D-peptide Design
D-peptides combine protease resistance with high target specificity, but computational design of D-peptide binders remains immature.
OpenAI-HuggingFace: A Reproduction & Lessons for Alignment Testing
In July
From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI
CoreWeave has built
Right Words, Wrong Moment: A Clinician-Grounded Analysis of Distress in 19,930 Conversations between Young People and ChatGPT
Young people increasingly turn to General-Purpose Conversational Agents (GPCAs), such as ChatGPT, in moments of distress.
Benchy: towards a universal language for task-oriented AI benchmarks
Benchy is a semantic language and execution engine for benchmarking AI programs.
T-RoPE: Time-Aware Rotary Position Embedding for Sequential Recommendation
Large-scale recommenders increasingly adopt the sequential generative recipe behind large language models, bringing the Transformer into recommendation along with design choices made for text, including Rotary Position Embedding (RoPE).
Audio LLMs Know When They Can’t Hear You
Audio large language models
Atelier: Learning Local Self-Supervised Features for CryoEM Volumes via Hypernetworks
CryoEM map interpretation requires features that are spatially localized, consistent across samples, and informative across spatial scales.
BioEVAL: A global, multi-institutional benchmark of large language and multimodal models for bioengineering
Large Language Models (LLMs) have demonstrated historic breakthroughs in general reasoning with early successes in biomedical science.
Do LLMs Understand Context? A Knowledge Graph-Based Evaluation Framework
While large language models (LLMs) have achieved remarkable linguistic capabilities, a profound question lingers at their core: do these models truly comprehend context or simply excel at pattern matching on an unprecedented scale?
CRC-Router: Risk-Constrained Routing for Medical Agentic AI Systems
Agentic AI systems are increasingly being explored in medical imaging to improve throughput and reduce clinician workload; however, safe deployment remains challenging because autonomous errors may propagate into downstream clinical decisions.
Thinking Less to Simulate Better: Intuitive Prompting Improves LLM Agents Simulating Individual Social Media Reactions, Including Unfamiliar Content
Platform policies are increasingly tested on artificial users, making agent fidelity important.
The Price of Thought: Does Test-Time Reasoning Pay in LLM Trading?
While inference-time reasoning in large language models (LLMs) promises better decision making, its higher computational cost may not yield better economic outcomes.
LLM Parkinsonism: Executive-Control Failure, Token-Inefficient Persistence, and an Uncertainty-Aware Global Executive Control Architecture for Autonomous Language-Model Agents
Large language models (LL
ScopeBench: Do Agents Preserve Engagement Boundaries Under Goal Pressure?
Agents are increasingly deployed with real autonomy in web application and network penetration testing, where a single out-of-scope action can breach a client’s engagement boundary.
Bridging LLM Agents and Data Spaces: An Architectural Mediation Approach using the Model Context Protocol
Data Spaces enable sovereign and governed data sharing across organizational boundaries, but their integration with AI agents remains challenging due to mismatches between probabilistic language model interactions and policy-driven data infrastructures.
Predicting Transmembrane Protein Topology from 3D Structure
This paper
Introducing Claude Sonnet 5.5
A clear upgrade over Sonnet 5 that runs 30% faster and costs up to 30% less for most work.
Bringing AI to Autonomous Systems – From Cognition to Collective Intelligence
The purpose of this article is to highlight the central role of autonomous systems as the ultimate stage in the development of AI, to explain the underlying technical challenges that require a combination of connectionist AI and symbolic AI, and to integrate AI and systems engineering.
Pretrained ASR Pseudo-labeling for Noisy Police Audio
Pretrained ASR systems perform poorly on noisy Broadcast Police Communication (BPC), hindering efforts to understand police decision-making.
When Is a Multi-Agent Code Judge Actually Grounded? Two Label-Free Measurements, and a Judge That Declines to Guess
When one language model judges whether another’s code is correct, it does not report the absence of evidence.
A Synthetic Ground-Truth Framework for the Evaluation of Explainable AI Methods
Evaluating explainable Artificial