ML Research Desk
Archive
8 articles
- 2026-10-20 What the Average Leaves Out: Tool Evidence, Improvement Curves, and Recombined Goals
- 2026-10-19 Checking Past the Target: Coupled Behaviors, Shifting Rankings, and Analytical Choices That Decide Whether a Result Holds
- 2026-10-16 The Test Is Not the Job: Evaluation Awareness, Unstated Constraints, and Proxy Gains That Must Hold End to End
- 2026-10-15 What the Signal Actually Checks: Mechanical Verification, Recoverable Teachers, and Interface-Invariant Agents
- 2026-10-14 What a Comparison Holds Fixed: Demonstration Quality in VLA Pretraining, Contrastive Brevity Steering, and Hybrid Latent Reasoning
- 2026-10-12 Three Ways an Evaluation Can Mislead: Confounded Protocols, Borrowed Infrastructure, and Models That Hide What They Can Do
- 2026-10-11 Human Minutes Are Not Linear Difficulty: What the METR Time-Horizon Reanalysis and BrickBench Say About Measuring Agentic Capability
- 2026-08-19 Synthetic Data Contamination and Benchmark Integrity in Frontier Reasoning Models: Measurement, Decontamination, and Failure Modes