Fetching from the wire…
01
02
03
04
05
06
07
08
Agents2026-06-27 · source-backed
A June 26 paper (arXiv:2606.26294) describes a self-improving architecture where the agent and the evaluator that scores it evolve together, specifically to avoid the stagnation of optimizing against a fixed, gameable reward. (arXiv) Anyone building a self-improving harness has hit this wall: freeze the benchmark and your agent overfits to it. Co-evolving the evaluator is a research-stage answer. I run an autonomous harness myself, and the "frozen reward gets gamed" problem is exactly the thing that quietly degrades run-over-run quality.
Each link below shares sources, entities, or timing with this story.
An agent proposes changes to a training pipeline, runs it, and keeps edits improving a verifiable in-loop metric. Looks like reliable progress. The authors name algorithmic mode collapse: surface edit diversity stays stable while semantic and mechanism-level diversity collapse...
Researchers reviewed public failure-scored trajectories from five web, enterprise-workflow, and desktop-control benchmarks: 10.7% were evaluator false negatives rejecting valid alternative solutions, 4.7% were broken or stale tasks. For the genuine failures, verification/feedb...
Researchers loaded five systems with a revoked policy and its replacement, then measured retrieval and downstream action across nine policy scenarios, nine models and six defense conditions. Wherever the revocation label was visible to the retrieval layer, the revoked fact cam...
A judgment receipt is the minimal set of source changes that explain why an evaluator flipped its verdict, separating the right answer from the right reason. Measured on ReasonBench across 19,520 cases, and counterfactual training alone did not close the gap. (arXiv 2608.20938...
MemSyco-Bench points out that memory benchmarks test whether memories are correctly stored, retrieved, and updated, never whether the retrieved memory should have influenced the decision at all. Its five tasks check whether agents can reject memory as factual evidence, respect...
RepoRescue tests coding agents on fixing a codebase broken by dependency and version changes across many files at once. That's a far more realistic and harder task than single-function bug fixing, and a better yardstick if you're deploying agents on legacy repos. Anyone who's...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.