En Passante №4

The through-line

The industry is consolidating around agentic patterns for production use. Cloud platforms (AWS, Cloudflare) are hardening agent deployment with governance and observability; local inference (NVIDIA) is gaining traction for latency-sensitive work; and developer tools (GitHub, Bedrock) are treating multi-agent orchestration as routine.

Issue signature

The Passing Shot

General AI

The largest brain map yet is a verification bottleneck, not an algorithm win

A complete male fruit fly connectome with 166,000 neurons and 125 million synapses shipped this week, and the constraint on the next one is human proofreading hours.

338 words · 1 source · verified

The Wire

Speculative Macro Commit for Faster Tool-Using Agents arXiv cs.AI
Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory arXiv cs.AI
A Prompt-Engineering Approach to Develop Scalable, Flexible, and Real-Time Hybrid Micro-Level Personalization in a General Purpose AI Teaching Assistant arXiv cs.AI
Caught in the Story: Narrative Captivity in Multi-turn LLMs Conversation arXiv cs.AI
Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection arXiv cs.AI
DuplexSpeechBench-IFEval: Evaluating Implicit Instruction Following in Full-Duplex Voice Agents arXiv cs.AI
Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents arXiv cs.AI
Beyond "Made with AI": Visualizing Provenance Density to Mitigate the Transparency Penalty arXiv cs.AI
AutoGraphForge: Towards Automated Graph Theory Discovery arXiv cs.AI
Making Every Tool Call Count: Necessary Tool-Evidence Path Rewards for Agentic Vision-Language Models arXiv cs.AI
GrowPage: On-Demand KV Budgeting for Efficient LLM Reasoning Serving arXiv cs.AI
PPO-STGNN: A Proximal Policy Optimization Approach with Spatio-Temporal Graph Neural Networks for DAG Task Scheduling in Cloud-Edge-End Computing arXiv cs.AI
What Matters for Aggressive Decoding-Time KV Eviction? Temporal Aggregation and Ranking Preservation arXiv cs.AI
CulturalMenuBench: Probing the Knowledge-Application Gap in Multimodal Culinary Reasoning arXiv cs.AI
NeoRed: A Knowledge-Logic-Alignment Multimodal Large Language Model for Neonatal Respiratory Disease Diagnosis arXiv cs.AI
Feature Reconfiguration With Visual Prior for Medical Lesion Segmentation arXiv cs.AI
Dalek: A Constructive Agent Machine arXiv cs.AI
GPS-Bench: A Governance Policy Benchmark for Automating Policy Analysis arXiv cs.AI
HalluPeer: A Taxonomy-driven Benchmark for Detecting Hallucinations in Scientific Peer Reviews arXiv cs.AI
The Attention Triangle in Audio-Video Models arXiv cs.AI
KC-Bench: A Dynamic Interactive Benchmark for Evaluating Knowledge Conflicts in LLM Agents arXiv cs.AI
A computable representation of the physical laboratory enables verifiable workflows arXiv cs.AI
Analysis of Prompt Engineering for Drug Toxicity Prediction arXiv cs.AI
Synthetic Semantic Supervision for Contrastive Code Representation Learning in Small Transformers: An Empirical Study arXiv cs.AI
Counterfactual Routing Using Integer Programming with Constraint Generation arXiv cs.AI