Deep Dives

Analysis

Deep-dive editorial on AI industry trends and original research

574 articles 48 pages

All Critical (9-10) Important (7-8) Notable (5-6) Logged (1-4) 574 articles

Editorial illustration for: GUARD-SLM: Token Activation-Based Defense Against Jailbreak Attacks for Small Language Models

GUARD-SLM: Token Activation Defense Targets Small Language Model Jailbreaks

4/10 4 min read 1 month ago

Editorial illustration for: ARTLAS: Mapping Art-Technology Institutions via Conceptual Axes, Text Embeddings, and Unsupervise

ARTLAS Maps 78 Cultural-Technology Institutions Using NLP Clustering

3/10 4 min read 1 month ago

Editorial illustration for: SkillTester: Benchmarking Utility and Security of Agent Skills

SkillTester Benchmarks AI Agent Skills on Utility and Security

4/10 4 min read 1 month ago

Editorial illustration for: The impact of multi-agent debate protocols on debate quality: a controlled case study

Debate Protocol Design Shapes Multi-Agent AI Outcomes, Controlled Study Finds

3/10 4 min read 1 month ago

Editorial illustration for: StepCache: Step-Level Reuse with Lightweight Verification and Selective Patching for LLM Serving

StepCache Cuts LLM Median Latency 99% via Step-Level Reuse

4/10 4 min read 1 month ago

Editorial illustration for: AI in Work-Based Learning: Understanding the Purposes and Effects of Intelligent Tools Among Stud

Philippine Intern Survey Maps Four Categories of AI Tool Use in OJT

3/10 4 min read 1 month ago

Editorial illustration for: The Last Fingerprint: How Markdown Training Shapes LLM Prose

Em Dashes in LLM Output Traced to Markdown Training Data

4/10 4 min read 1 month ago

Editorial illustration for: ScoringBench: A Benchmark for Evaluating Tabular Foundation Models with Proper Scoring Rules

ScoringBench Ranks Tabular AI Models on Full Distribution Accuracy

3/10 4 min read 1 month ago

Editorial illustration for: Uncertainty Gating for Cost-Aware Explainable Artificial Intelligence

Epistemic Uncertainty Proposed as Routing Signal for Cheaper, More Reliable AI Explanations

3/10 4 min read 1 month ago

Editorial illustration for: ATP-Bench: Towards Agentic Tool Planning for MLLM Interleaved Generation

ATP-Bench: Researchers Benchmark 10 MLLMs on Agentic Tool Planning

3/10 4 min read 1 month ago

Editorial illustration for: ShapE-GRPO: Shapley-Enhanced Reward Allocation for Multi-Candidate LLM Training

ShapE-GRPO Uses Shapley Values to Fix GRPO Free-Rider Problem in LLM Training

3/10 4 min read 1 month ago

Editorial illustration for: Tracking vs. Deciding: The Dual-Capability Bottleneck in Searchless Chess Transformers

Dual-Capability Bottleneck in Chess AI Formalized, Model Hits Lichess 2570

3/10 4 min read 1 month ago

← Prev 1 … 32 33 34 35 36 … 48 Next → Page 34 of 48

📬 Get AI news daily → Subscribe Free