HELIX: Model-Harness Co-evolution for Recursive Self-Improvement

• URL: http://arxiv.org/abs/2608.13951v1
• Date: Fri, 14 Aug 2026 04:44:53 GMT
• Title: HELIX: Model-Harness Co-evolution for Recursive Self-Improvement
• Abstract summary: We present HELIX, a source-traceable substrate for harness evolution.<n>We show how harness evolution expands current capability and creates learning signal for the next model.
• Score: 8.484282837156933
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Scaling agent capability has largely focused on improving the model, yet an interactive agent acts through a runtime harness that mediates context, tools, control flow, and stopping. The harness shapes both what a model can accomplish and the trajectories from which it learns. This coupling motivates model-harness co-evolution for recursive self-improvement: build harnesses for a fixed model, update the model from verified sibling trajectories, and rebuild the harnesses as model capabilities change. Realizing this loop requires a controlled way to evolve harnesses while preserving intervention identity and effect. We present HELIX, a source-traceable substrate for harness evolution. HELIX decomposes agent systems into typed ports, reusable atoms, recipes, product shells, and runtime policies. It makes interventions explicit and auditable while retaining trajectories, test outcomes, and provenance. Harness evolution thus serves two linked roles: improving fixed-model execution and producing matched successes, regressions, near misses, and alternative solutions as data for subsequent model improvement. We evaluate HELIX in one evolution round on code repair. A 65-candidate portfolio discovers a fixed harness that improves task coverage by 4.0% over Pi, while the full portfolio exposes up to 58.0% more verified coverage through complementary sibling behavior. Selected candidates are assessed with repeated runs and the SWE-bench evaluator. A 200-slot sibling slice yields 438 verified SFT, critic, filter, and preference records. These results show how harness, model, and data form a feedback system: harness evolution expands current capability and creates learning signal for the next model; model updates motivate the next round of harness evolution. HELIX provides an auditable interface for studying this recursive process. Code is available at https://github.com/HKUDS/HELIX.

Related papers

• Co-Evolving Harnesses and Models: On-Policy Correction Helps Weaker Models Catch Up Where Imitation Fails [23.287109132562403]
  Automated harness evolution can enable smaller models to perform well on domain-specific tasks at a fraction of frontier-model cost.<n>We show that imitation transfers knowledge and increases scaffold usage, but disrupts model-harness fit.<n>We develop an on-policy expert-correction pipeline, automated by a meta-level MLE agent, that localizes the failing turn in the weaker model's own rollout and asks the expert to rewrite only that turn.
  arXiv  Detail & Related papers  (2026-09-08T17:53:49Z)
• HarnessDev: Can LLMs Create and Evolve Their Own Agent Harness? [61.02699867449589]
  We introduce HarnessDev, a benchmark that shifts the unit of evaluation from task outputs to run infrastructure.<n>In Creation, the agent starts from a minimal seed and a small number of cases, then builds a complete execution system.<n>In Evolution, it starts from its own created harness and iteratively revises it using downstream execution feedback, with the goal of improving benchmark performance.
  arXiv  Detail & Related papers  (2026-09-01T15:45:33Z)
• Harness Continual Learning: Continual Adaptation Beyond Model Parameters [9.54786133734043]
  We formulate a new continual learning paradigm in which the harness evolves around a frozen foundation model.<n>A candidate commits a harness only after checking current improvement, historical retention, and validity.<n>Experiments demonstrate capability accumulation and failure recovery, with relative gains exceeding 10% over corresponding baselines in multiple settings.
  arXiv  Detail & Related papers  (2026-08-19T15:12:31Z)
• One Recipe, Many Harnesses: What Self-Evolution Encodes Across Languages and Models [46.83445434865936]
  Self-evolving harnesses are closed-loop systems in which an agent inspects its own rollouts and edits its prompts, tools, and memory.<n>It remains unclear whether they encode benchmark-specific adaptations, language-specific engineering knowledge, or compensation for limitations of the underlying model.
  arXiv  Detail & Related papers  (2026-08-10T19:45:45Z)
• EvolveNet: Collaborative Harness Evolution for Agent Self-Improvement [70.68218215070745]
  Existing approaches assume that all execution experience can be routed to a single harness.<n>We introduce EvolveNet, a paradigm of harness evolution that moves experience extraction to the data.
  arXiv  Detail & Related papers  (2026-08-05T15:37:18Z)
• Automated Discovery Has No Universally Superior Harness [63.53666712024897]
  We decompose OpenEvolve-style evolutionary search and the TTT-Discover search harness into its constituent components.<n>We evaluate 30 budget-matched harnesses across 12 model-problem pairs.
  arXiv  Detail & Related papers  (2026-07-20T17:59:37Z)
• Rethinking the Evaluation of Harness Evolution for Agents [83.07258924910069]
  Existing harness evolution methods use unit test cases to search for harness configurations and then report final performance on the same public benchmark.<n>This protocol raises two fundamental concerns.<n>As in agentic test-time scaling, it should therefore be compared with simple task-level search baselines under matched feedback and inference budgets.
  arXiv  Detail & Related papers  (2026-07-14T00:18:42Z)
• HarnessX: A Composable, Adaptive, and Evolvable Agent Harness Foundry [35.87794858139959]
  HarnessX is a foundry for composable, adaptive, and evolvable agent harnesses.<n>It assembles typed harness primitives via a substitution algebra, adapts them through AEGIS.<n>It closes the harness-model loop by turning trajectories into both harness updates and model training signal.
  arXiv  Detail & Related papers  (2026-06-12T08:27:11Z)
• Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents [82.27610290890475]
  We analyze two harness self-evolution capabilities: (i) harness-updating, the capability to produce useful persistent harness updates from execution evidence; (ii) harness-benefit, the capability to benefit from updated harnesses during task solving.<n>Our analysis reveals two findings. First, harness-updating is flat in base capability: models from different capability tiers produce harness updates that lead to surprisingly similar gains.<n>Second, harness-benefit is non-monotonic in base capability: weak-tier models benefit little from updated harnesses, mid-tier models benefit most, and strong-tier models benefit less than mid-tier.
  arXiv  Detail & Related papers  (2026-05-28T22:16:14Z)
• DemoEvolve: Overcoming Sparse Feedback in Agentic Harness Evolution with Demonstrations [12.048713464721665]
  DemoEvolve is a demonstration-bootstrapped approach to harness evolution.<n>We study this paradigm as a form of sample-efficient fast adaptation.<n>We show that DemoEvolve produces more effective and auditable harness edits.
  arXiv  Detail & Related papers  (2026-05-23T12:10:52Z)
• Provably Efficient Online RLHF with One-Pass Reward Modeling [70.82499103200402]
  Reinforcement Learning from Human Feedback has shown remarkable success in aligning Large Language Models with human preferences.<n>Online RLHF has emerged as a promising direction, enabling iterative data collection and refinement.<n>We propose a one-pass reward modeling method that eliminates the need to store historical data and achieves constant-time updates per iteration.
  arXiv  Detail & Related papers  (2025-02-11T02:36:01Z)
• AvgOut: A Simple Output-Probability Measure to Eliminate Dull Responses [97.50616524350123]
  We build dialogue models that are dynamically aware of what utterances or tokens are dull without any feature-engineering. The first model, MinAvgOut, directly maximizes the diversity score through the output distributions of each batch. The second model, Label Fine-Tuning (LFT), prepends to the source sequence a label continuously scaled by the diversity score to control the diversity level. The third model, RL, adopts Reinforcement Learning and treats the diversity score as a reward signal.
  arXiv  Detail & Related papers  (2020-01-15T18:32:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.