AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses

• URL: http://arxiv.org/abs/2608.12307v1
• Date: Wed, 12 Aug 2026 17:53:18 GMT
• Title: AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses
• Abstract summary: We study whether a stronger builder model can construct inference-time harnesses that help a weaker target model solve tasks more reliably without any parameter updates.<n> Empirically, this form of test-time capability transfer is highly effective, nearly doubling average target-model performance from 0.49 to 0.91.<n>We find that builder-model reasoning effort improves harness quality monotonically, platform effects are modest relative to the builder model's own capability, and weaker target models receive the largest gains.
• Score: 72.0777620499307
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent work on distillation transfers the capabilities of large models to smaller ones often by updating the latter's parameters, through teacher forcing, on-policy distillation, and related training-time methods. In this paper, we ask whether such transfer can instead occur at test time. We study strong-to-weak scaffolding: whether a stronger builder model can construct inference-time harnesses that help a weaker target model solve tasks more reliably without any parameter updates. Using four representative Theory-of-Mind benchmarks, each builder model uses 5% of the data as a validation set to iteratively refine its harness over multiple rounds, after which the finalized harness is evaluated on the full test set. Empirically, this form of test-time capability transfer is highly effective, nearly doubling average target-model performance from 0.49 to 0.91. Our analysis shows that the gains come primarily from offloading unstable model reasoning into deterministic code, benchmark-specific routing, and strict answer-format enforcement, rather than from encouraging the target model to reason more extensively or sample more broadly. We further find that builder-model reasoning effort improves harness quality monotonically, platform effects are modest relative to the builder model's own capability, and weaker target models receive the largest gains. These results suggest that inference-time harness design is an important complement to conventional training-time distillation, enabling strong models to transfer cognitive structure to weaker models without retraining.

Related papers

• FABSVer: Faster Training and Better Self-Verification for LLM Mathematical Reasoning [8.96133612398978]
  Existing approaches that equip models with self-verification typically treat solution generation and verification as two separate tasks.<n>We propose FABSVer, which fuses these two tasks into a single generation pass.<n>We show that FABSVer achieves superior self-verification and reasoning performance across three model scales.
  arXiv  Detail & Related papers  (2026-05-27T12:26:34Z)
• On Predicting the Post-training Potential of Pre-trained LLMs [60.07459271263409]
  We introduce a new task of predicting post-training potential - forecasting a base model's performance before post-training.<n>We propose RuDE, a unified framework that bypasses the generation gap of base models by leveraging response discrimination.<n>Experiments demonstrate a correlation greater than 90% with post-training performance.
  arXiv  Detail & Related papers  (2026-05-12T11:33:49Z)
• Model Whisper: Steering Vectors Unlock Large Language Models' Potential in Test-time [6.741914038966904]
  We introduce a lightweight component, Test-Time Steering Vectors (TTSV), which is prepended to the input while keeping the model's parameters entirely frozen.<n>TTSV is both lightweight and highly efficient to optimize, making it a true plug-and-play enhancement.<n>Our approach exhibits robust generalization, with its steering vectors proving highly transferable across diverse tasks.
  arXiv  Detail & Related papers  (2025-12-04T12:36:16Z)
• Self-Improving LLM Agents at Test-Time [49.9396634315896]
  One paradigm of language model (LM) fine-tuning relies on creating large training datasets.<n>In practice, gathering large sets of data is inefficient, and training on them is prohibitively expensive.<n>We study two variants of this approach: Test-Time Self-Improvement (TT-SI) and Test-Time Distillation (TT-D)
  arXiv  Detail & Related papers  (2025-10-09T06:37:35Z)
• Leveraging Reasoning Model Answers to Enhance Non-Reasoning Model Capability [16.441081996257576]
  We propose leveraging reasoning-intensive models to improve less computationally demanding, non-reasoning models.<n>We demonstrate consistent improvements across various benchmarks, underscoring the potential of this approach for advancing the ability of models to answer questions directly.
  arXiv  Detail & Related papers  (2025-04-13T16:26:56Z)
• On the Diminishing Returns of Complex Robust RAG Training in the Era of Powerful LLMs [85.688901949146]
  We investigate the question: does the benefit of complex robust training methods diminish as language models become more powerful?<n>Our analysis reveals a consistent trend: emphthe marginal robustness benefit of sophisticated training strategies decreases substantially as model capacity increases.<n>Further investigation demonstrates that stronger models naturally exhibit better confidence calibration, cross-dataset generalization capability, and more effective attention patterns, even under simple training regimes.
  arXiv  Detail & Related papers  (2025-02-17T03:34:31Z)
• Large Pre-Training Datasets Don't Always Guarantee Robustness after Fine-Tuning [29.69990405081772]
  We propose the Robustness Inheritance Benchmark (ImageNet-RIB) to assess robustness preservation in fine-tuned models.<n>We find that fine-tuning reduces robustness across pretrained models.<n>Models pretrained on the largest and most diverse datasets exhibit both larger robustness losses and lower absolute robustness after fine-tuning on small datasets.
  arXiv  Detail & Related papers  (2024-10-28T22:33:22Z)
• A Cost-Aware Approach to Adversarial Robustness in Neural Networks [1.622320874892682]
  We propose using accelerated failure time models to measure the effect of hardware choice, batch size, number of epochs, and test-set accuracy. We evaluate several GPU types and use the Tree Parzen Estimator to maximize model robustness and minimize model run-time simultaneously.
  arXiv  Detail & Related papers  (2024-09-11T20:43:59Z)
• QualEval: Qualitative Evaluation for Model Improvement [82.73561470966658]
  We propose QualEval, which augments quantitative scalar metrics with automated qualitative evaluation as a vehicle for model improvement. QualEval uses a powerful LLM reasoner and our novel flexible linear programming solver to generate human-readable insights. We demonstrate that leveraging its insights, for example, improves the absolute performance of the Llama 2 model by up to 15% points relative.
  arXiv  Detail & Related papers  (2023-11-06T00:21:44Z)
• Voting based ensemble improves robustness of defensive models [82.70303474487105]
  We study whether it is possible to create an ensemble to further improve robustness. By ensembling several state-of-the-art pre-trained defense models, our method can achieve a 59.8% robust accuracy.
  arXiv  Detail & Related papers  (2020-11-28T00:08:45Z)

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.