The View from Nowhere: Population-Centered Agent Evolution
Shuo Yan, Ningning Hou, Haoyu He, Haochen Pei and Shang Gao (Macquarie University, Monash University) show that agent-evolution runs solve many tasks somewhere in the population that the best agent never solves, and propose Population-Centered Evolution, which feeds population-wide evidence into each parent's harness update.
Ask this paper
The gap. Population coverage (tasks solved by at least one variant) is well above the best single agent's score. Capabilities that appear across the population are not consolidated into the leaders.
Method. PCE distills task records and modification histories from the whole population into a population block of recurring failures, complementary successes and the effects of past modifications. That block guides edits to a selected parent's harness.
Results. Under matched models, initial harnesses and schedules, PCE reaches 72.1% on SWE-bench Verified and 96.7% on Polyglot, 3.2 and 5.9 points above a strong group-evolution baseline.
Gap closed. PCE cuts the gap between population coverage and best-agent performance by one-third on SWE-bench and closes it on Polyglot.
Abstract
Self-improving AI systems can revise the designs that govern their own behavior. Yet existing approaches center improvement on selected parent agents and their evolutionary experience, leaving the broader population largely underused. We identify a substantial gap between population coverage (i.e., the tasks solved by at least one evolved variant) and the performance of the best individual agent. This gap shows that useful capabilities and diagnostic evidence emerge across the population without being consolidated into its leading agents. These observations motivate the view from nowhere, a principle for agent evolution: an agent should be improved by interpreting its behavior through population-wide patterns of success, failure, and prior self-modification. We instantiate this principle as Population-Centered Evolution (PCE). PCE distils task records and modification histories from the population into a population block, which captures recurring failures, complementary successes, and the effects of previously attempted modifications to guide improvements to a selected parent's harness. Under matched models, initial harnesses, and evolution schedules, PCE achieves 72.1% success on SWE-bench Verified and 96.7% on Polyglot, improving over a strong group-evolution baseline by 3.2 and 5.9 percentage points. In our evaluation, PCE reduces the gap between population coverage and best-agent performance by one-third on SWE-bench and closes it on Polyglot. These results suggest that treating the population as evidence--not merely as candidates for selection--provides a complementary principle for agent evolution.