Tag
NLD-Uni 提出一个统一的离散扩散大语言模型，通过跨范式在策略蒸馏和渐进式自我精炼扩散，将自回归 VLM 的视觉理解能力与连续流模型的视觉生成能力统一起来，在多项视觉理解与生成任务上达到统一 DLLM 中的最先进性能。
This article describes Sebastian Raschka's YouTube video episode on reasoning techniques in AI, focusing on log-probability scoring and self-refinement to improve model performance without retraining.
The paper introduces DENSE, a method for distilling AI agent execution traces into evidence-grounded shortcut trees for self-refinement, achieving improved performance on Terminal-Bench without post-hoc outcome labels.
This paper introduces Chain-of-Experience (CoE), a framework for continual LLM improvement through iterative test-time interactions with self and environmental feedback, showing consistent gains and lower API costs across multiple domains and models.
A comprehensive survey of 1,250 papers (2024–2026) on recursive self-improvement in AI, proposing a taxonomy distinguishing bounded self-refinement from open-ended recursive self-improvement, and analyzing the evaluator design space and failure modes.
A new method, FLORA (Frequency-corrected Learning of Ordered Rank Alignment), improves LLMs by aligning generator and validator modes using a principled frequency correction. Experiments show substantial gains in G-V consistency and generator performance on benchmarks like IFEval and HumanEval.
This paper investigates whether natural-language feedback leads to improvement beyond repeated attempts alone in multi-turn language agent settings. Using a controlled student-teacher protocol across multiple benchmarks, the authors find that self-generated feedback adds little, while strong external teachers yield larger gains, and that the student's ability to act on feedback is a key bottleneck.
Flow Reasoning Models (FRMs) introduce a training and test-time-scaling framework for discrete flow models on structured reasoning tasks. By using self-verification and self-conditioning, FRMs achieve nearly 100% solve rates on Sudoku and Zebra puzzles with far fewer passes than previous baselines.
This paper proposes a symbolic feedback-driven iterative self-refinement framework to improve the robustness and reliability of large language models in long-horizon planning tasks. The method uses natural language prompting, a symbolic verifier, and a plan recognizer to enhance feasibility and correctness.
This paper introduces RePro, a framework that trains LLM agents to self-generate progress signals through a forward-then-reflect rollout paradigm, achieving up to 12% absolute success rate gains on WebShop, ALFWorld, and Sokoban benchmarks.
AutoMine is a robust self-refining scenario mining method using LLMs and VLMs to mine high-value scenarios from autonomous driving logs, achieving top scores in the Argoverse 2 Scenario Mining Competition at CVPR 2026.