The paper introduces DOMINO, a novel inductive framework that synthesizes domain-specific data for LLMs using only reference examples, significantly improving performance on challenging, implicitly de…

The paper introduces State-Adaptive Prompt Optimization (SAPO), a novel training strategy that treats prompts as dynamic variables to achieve robust fine-tuning, significantly mitigating catastrophic…

The paper proposes EKSFT, a selective fine-tuning method that masks high-entropy or high-KL divergence tokens during Supervised Fine-Tuning (SFT) to prevent distribution shift and improve subsequent R…

This paper introduces Anchored Weight Decay (AWD), a regularization technique that effectively prevents prior-task forgetting during LLM fine-tuning with Evolution Strategies (ES), positioning ES as a…

The paper introduces Weak-Critic Strong Oversight, a method where a weak model guides a strong model's self-improvement by providing non-misleading revision directions, leading to scalable oversight.

The paper introduces Temperature-Scaled On-Policy Self-Distillation (TS-OPSD), a novel method that internalizes temperature-based policy reheating into model parameters to combat entropy collapse in r…

The paper introduces FoLoRA, a novel optimization framework that uses a generalized Rayleigh quotient to achieve a superior balance between adapting foundation models to specific tasks and preserving…

The paper introduces Sparse Memory-Efficient Training (SMET), a method that stabilizes and optimizes Dynamic Sparse Training (DST) for large language models, enabling stable and memory-efficient spars…

The paper proposes a local perturbation theory showing that cross-domain interference in multi-domain RL occurs via a low-dimensional shared conflict subspace, which can be selectively mitigated by sh…

DASH introduces a dual-branch distillation framework to effectively compress class-conditional diffusion models by independently supervising both score branches, significantly preserving guidance fide…

The paper proposes Distribution-Aligned Self-Distillation (DASD) to improve self-distillation by dynamically filtering high-perplexity tokens, thereby preserving useful logical knowledge while suppres…

The paper introduces RefusalGuard, a novel fine-tuning framework that preserves the geometric structure of safety-relevant representations in LLMs, thereby mitigating the degradation of refusal behavi…

TRACE proposes a novel method to mitigate catastrophic forgetting in continual LLM fine-tuning by identifying and isolating a small, task-specific subset of essential parameters for each task.

The paper proposes Visual Gradient Steering (VGS), a method that decomposes the distillation loss into language and visual components and steers the optimization to prioritize visual grounding, signif…

The paper introduces LearnWeak, an annotation-free framework that automatically specializes small computer-use agents by identifying and targeting their specific weaknesses using a stronger reference…

The paper investigates how dynamic adversarial fine-tuning (R2D2) reorganizes the internal mechanisms (refusal geometry) of safety-aligned language models, finding that it shifts the optimal refusal c…

The paper introduces RA-MoE, a novel fine-tuning framework that leverages the internal routing structure of Mixture-of-Experts (MoE) models to improve performance on multilingual downstream tasks by a…

The paper introduces Involuntary In-Context Learning (IICL), an effective few-shot pattern completion attack that can bypass safety alignments in large language models, achieving a 24.0% bypass rate a…