SIA: Self Improving AI with Harness & Weight Updates

• URL: http://arxiv.org/abs/2605.27276v2
• Date: Thu, 28 May 2026 08:21:29 GMT
• Title: SIA: Self Improving AI with Harness & Weight Updates
• Abstract summary: Humans are the bottleneck in building and improving AI.<n>We propose SIA, a self-improving loop in which a language-model agent updates both the harness and the weights of a task-specific agent.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-sa/4.0/
• Abstract: Humans are the bottleneck in building and improving AI. Both the models and the agents that wrap them are written, tuned, and corrected by people. The long-horizon goal of an AI that can figure out how to improve itself remains open. Two largely disjoint research lines attack this bottleneck. The harness-update school has a meta-agent rewrite the scaffold of a task-specific agent (its tools, prompts, retry logic, and search procedure) while the model weights are held fixed. The test-time training school uses hand-written RL pipelines to update the model's own weights on task feedback while the harness is held fixed. These two silos operate in isolation. We propose SIA, a self-improving loop in which a language-model agent (the Feedback-Agent) updates both the harness and the weights of a task-specific agent. We evaluate across three contrasting domains: Chinese legal charge classification, low-level GPU kernel optimisation, and single-cell RNA denoising. Combining both levers outperforms scaffold iteration alone on all three benchmarks. SIA-W+H achieves 25.1% over prior SOTA on LawBench, 12.4% faster GPU kernels than prior SOTA (1,017 vs 1,161 μs), and 20.4% over prior SOTA on denoising. Harness updates make the model agentic, shaping how it searches and acts, while weight updates build the domain intuition that no prompt or scaffold can instil.

Related papers

• WHALE: A Simple Recipe for Joint Harness-Weight Optimization [48.550149284101536]
  Weight-Harness Alternating LEarning (WHALE) is a recipe that alternates two phases: updating the model under the current harness, then searching for a better harness under the updated model.<n>Using Qwen3.5-2B/4B agents across three domains (search question answering, mathematical reasoning, and chess puzzles), WHALE outperforms weight-only, harness-only, and Fast-Slow Training by 4.15-24.38 percentage points in best mean@8 accuracy.
  arXiv  Detail & Related papers  (2026-08-31T18:12:55Z)
• Serving Masked Diffusion LLMs: Characterization and Design Principles from Real Hardware [1.7286026533711227]
  Masked diffusion language models (dLLMs) can generate text faster than autoregressive (AR) models, since they denoise many tokens at once.<n>Recent systems have begun building serving infrastructure for dLLMs, but none first measure how these models behave under real, concurrent serving.<n>We characterize dLLM serving to close this gap, using LLaDA-8B-Instruct with a D2F (Discrete Diffusion Forcing) LoRA adapter on a single NVIDIA H200 GPU.
  arXiv  Detail & Related papers  (2026-08-24T20:16:48Z)
• 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)
• From Model Scaling to System Scaling: Scaling the Harness in Agentic AI [4.802305157491253]
  This paper studies the next major bottleneck in agentic AI as system scaling.<n>We refer to this shift as scaling the harness, treating the structured execution layer around a foundation model as a first-class object of design, evaluation, and optimization.<n>Our main claim is that future progress in agentic AI will depend as much on system design as on stronger foundation models.
  arXiv  Detail & Related papers  (2026-05-25T17:59:36Z)
• Workspace Optimization: How to Train Your Agent [49.579258311548635]
  We argue it is the agent's emphworkspace, the structured external substrate it reads, writes, and tests.<n>We propose a principled way to evolve the workspace, mirroring the structure of weight-space training.<n>We instantiate the idea in DreamTeam, a multi-agent harness for ARC-AGI-3.
  arXiv  Detail & Related papers  (2026-05-10T16:52:10Z)
• ClawArena: Benchmarking AI Agents in Evolving Information Environments [61.664633997138004]
  ClawArena is a benchmark for evaluating AI agents in evolving information environments.<n>Each scenario maintains a complete hidden ground truth while exposing the agent only to noisy, partial, and sometimes contradictory traces.<n> evaluation is organized around three coupled challenges: multi-source conflict reasoning, dynamic belief revision, and implicit personalization.
  arXiv  Detail & Related papers  (2026-04-05T17:55:23Z)
• Holistic Agent Leaderboard: The Missing Infrastructure for AI Agent Evaluation [87.47155146067962]
  We provide a standardized evaluation harness that orchestrates parallel evaluations across hundreds of tasks.<n>We conduct three-dimensional analysis spanning models, scaffolds, and benchmarks.<n>Our analysis reveals surprising insights, such as higher reasoning effort reducing accuracy in the majority of runs.
  arXiv  Detail & Related papers  (2025-10-13T22:22:28Z)
• Reinforcement Learning for Machine Learning Engineering Agents [52.03168614623642]
  We show that agents backed by weaker models that improve via reinforcement learning can outperform agents backed by much larger, but static models.<n>We propose duration- aware gradient updates in a distributed asynchronous RL framework to amplify high-cost but high-reward actions.<n>We also propose environment instrumentation to offer partial credit, distinguishing almost-correct programs from those that fail early.
  arXiv  Detail & Related papers  (2025-09-01T18:04:10Z)
• Agentic Program Repair from Test Failures at Scale: A Neuro-symbolic approach with static analysis and test execution feedback [11.070932612938154]
  We develop an Engineering Agent that fixes the source code based on test failures at scale across diverse software offerings.<n>We provide feedback to the agent through static analysis and test failures so it can refine its solution.<n>In a three month period, 80% of the generated fixes were reviewed, of which 31.5% were landed.
  arXiv  Detail & Related papers  (2025-07-24T19:12:32Z)
• NoHumansRequired: Autonomous High-Quality Image Editing Triplet Mining [34.41455335920275]
  We present an automated, modular pipeline that mines high-fidelity triplets across domains, resolutions, instruction complexities, and styles.<n>Inversion and compositional bootstrapping enlarge the mined set by approx. 2.6x, enabling large-scale high-fidelity training data.<n>We release NHR-Edit, an open dataset of 720k high-quality triplets, curated at industrial scale via millions of guided generations and validator passes.
  arXiv  Detail & Related papers  (2025-07-18T17:50:00Z)
• ABBA-Adapters: Efficient and Expressive Fine-Tuning of Foundation Models [10.17362679822278]
  Large Language Models have demonstrated strong performance across a wide range of tasks, but adapting them efficiently to new domains remains a key challenge.<n>We introduce ABBA, a new PEFT architecture that re parameterizes the update as a Hadamard product of two independently learnable low-rank matrices.<n>In contrast to prior work, ABBA fully decouples the update from the pre-trained weights, enabling both components to be optimized freely.
  arXiv  Detail & Related papers  (2025-05-20T11:43:25Z)
• How Well Can AI Build SD Models? [0.0]
  We introduce two metrics for evaluation of AI-generated causal maps: technical correctness (causal translation) and adherence to instructions (conformance)<n>We tested 11 different LLMs on their ability to do causal translation as well as conform to user instruction.
  arXiv  Detail & Related papers  (2025-03-19T14:48:47Z)
• Model Ratatouille: Recycling Diverse Models for Out-of-Distribution Generalization [99.6826401545377]
  Foundation models are redefining how AI systems are built. Practitioners now follow a standard procedure to build their machine learning solutions. We propose model ratatouille, a new strategy to recycle the multiple fine-tunings of the same foundation model on diverse auxiliary tasks.
  arXiv  Detail & Related papers  (2022-12-20T17:21:46Z)
• Retrieval-Augmented Reinforcement Learning [63.32076191982944]
  We train a network to map a dataset of past experiences to optimal behavior. The retrieval process is trained to retrieve information from the dataset that may be useful in the current context. We show that retrieval-augmented R2D2 learns significantly faster than the baseline R2D2 agent and achieves higher scores.
  arXiv  Detail & Related papers  (2022-02-17T02:44:05Z)
• Improving Neural Machine Translation by Denoising Training [95.96569884410137]
  We present a simple and effective pretraining strategy Denoising Training DoT for neural machine translation. We update the model parameters with source- and target-side denoising tasks at the early stage and then tune the model normally. Experiments show DoT consistently improves the neural machine translation performance across 12 bilingual and 16 multilingual directions.
  arXiv  Detail & Related papers  (2022-01-19T00:11:38Z)
• The USYD-JD Speech Translation System for IWSLT 2021 [85.64797317290349]
  This paper describes the University of Sydney& JD's joint submission of the IWSLT 2021 low resource speech translation task. We trained our models with the officially provided ASR and MT datasets. To achieve better translation performance, we explored the most recent effective strategies, including back translation, knowledge distillation, multi-feature reranking and transductive finetuning.
  arXiv  Detail & Related papers  (2021-07-24T09:53:34Z)

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.