arXiv subjects

Jingang Zhou

Publications and source records attributed to Jingang Zhou.

5 recordsLinked to original sources

arXiv2026

CAMFT: Conflict-Aware Mergeable Fine-Tuning for Large Language Models

Model merging has emerged as a promising paradigm for integrating multiple task-specific capabilities into a single large language model. However, existing methods predominantly focus on post-hoc processing of independently fine-tuned models, overlooking how the training phase itself impacts cross-task compatibility. Resolving parameter conflicts after fine-tuning is inherently sub-optimal. To address this, we propose CAMFT, a Conflict-Aware Mergeable Fine-Tuning method that makes task adaptation both efficient and mergeaware. CAMFT treats mergeability as a property shaped during fine-tuning, rather than only a problem to be solved after fine-tuning. By guiding each task to update sparse coordinates with lower cross-task conflict, CAMFT produces task updates that are efficient to train and more compatible for downstream model merging. Extensive experiments demonstrate that CAMFT outperforms standard finetuning baselines in multi-task merging scenarios. Codes are available at https://github.com/gyanchow/CAMFT-LLM.

cs.LG↗

arXiv2026

Teacher Should Think Ahead: Adaptive Continuations for Reliable On-Policy Distillation

On-policy distillation (OPD) is a promising approach for transferring knowledge between language models, where a student receives dense token-level supervision along its own generated trajectories. However, teacher supervision can be unreliable when conditioned on incomplete or low-quality student prefixes. We identify Teacher Uncertainty Contraction (TUC), a systematic phenomenon whereby the teacher's predictive uncertainty decreases as it continues from a student-generated prefix. We theoretically characterize this trade-off through a variance-bias decomposition of teacher-branch gradients, showing that uncertainty contraction reduces variance while teacher-student path divergence increases bias, thereby favoring a finite continuation. Guided by this insight, we propose Adaptive-Continuations On-Policy Distillation (AC-OPD), which augments informative states along student rollouts with teacher continuations and adaptively selects their effective supervision horizons. Experiments on mathematical reasoning and code generation across model scales demonstrate that AC-OPD consistently improves over standard OPD. Controlled-continuations and matched-budget analyses further validate the adaptive-continuations design, highlighting adaptive teacher continuations as an effective principle for reliable on-policy distillation.The code will be made publicly available upon publication.

cs.LG↗

arXiv2026

APTER: Adaptive Post-Training with Expert-Grounded Rubrics

As large language models enter professional domains, they must satisfy domain constraints, include critical evidence, and provide complete reasoning rather than merely produce fluent responses. Existing post-training methods often rely on holistic preferences or outcome-level verification, while recent rubric-based methods usually generate rubrics independently for each query. In specialized domains, such unconstrained rubrics may omit critical requirements and vary across samples, hindering the diagnosis and targeted repair of persistent capability deficiencies. We propose APTER (Adaptive Post-Training with Expert-Grounded Rubrics), a framework that integrates structured domain knowledge into fine-grained evaluation, optimization, and diagnosis for specialized complex reasoning. First, expert-grounded rubric construction starts from an expert criteria framework built by domain experts, where each criterion represents a stable professional capability. For each query, APTER selects relevant criteria and instantiates them into query-level rubrics linked to their source criteria, turning reusable expert criteria into executable query-level supervision without reference answers. Second, adaptive post-training uses rubric verdicts as both optimization and criterion-level diagnostic signals. Aggregating low-scoring verdicts by criterion ID reveals persistent deficiencies and triggers targeted supervised fine-tuning updates during reinforcement learning. Experiments on mathematical reasoning and medical question answering show consistent gains across both domains. Across three model generations, APTER improves the mathematics and medical averages over the corresponding base models by up to 15.86 and 8.04 points, respectively. Code and rubric datasets are available at https://github.com/AntDT-APTER/APTER.

cs.AI↗

arXiv2025

Continual Learning for Generative AI: From LLMs to MLLMs and Beyond

The rapid advancement of generative models has empowered modern AI systems to comprehend and produce highly sophisticated content, even achieving human-level performance in specific domains. However, these models are fundamentally constrained by \emph{catastrophic forgetting}, \ie~a persistent challenge where models experience performance degradation on previously learned tasks when adapting to new tasks. To address this practical limitation, numerous approaches have been proposed to enhance the adaptability and scalability of generative AI in real-world applications. In this work, we present a comprehensive survey of continual learning methods for mainstream generative AI models, encompassing large language models, multimodal large language models, vision-language-action models, and diffusion models. Drawing inspiration from the memory mechanisms of the human brain, we systematically categorize these approaches into three paradigms: architecture-based, regularization-based, and replay-based methods, while elucidating their underlying methodologies and motivations. We further analyze continual learning setups for different generative models, including training objectives, benchmarks, and core backbones, thereby providing deeper insights into the field. The project page of this paper is available at https://github.com/Ghy0501/Awesome-Continual-Learning-in-Generative-Models.

cs.LG↗

arXiv2020

Enhanced anisotropic superconductivity in the topological nodal-line semimetal InxTaS2

Coexistence of topological bands and charge density wave (CDW) in topological materials has attracted immense attentions because of their fantastic properties, such as axionic-CDW, three-dimensional quantum Hall effect, etc. In this work, a nodal-line semimetal InxTaS2 characterized by CDW and superconductivity is successfully synthesized, whose structure and topological bands (two separated Wely rings) are similar to In0.58TaSe2. A 2 x 2 commensurate CDW is observed at low temperature in InxTaS2, identified by transport properties and STM measurements. Moreover, superconductivity emerges below 0.69 K, and the anisotropy ratio of upper critical field [Gamma = H||ab c2(0)=H||c c2(0)] is significantly enhanced compared to 2H-TaS2, which shares the same essential layer unit. According to the Lawrence-Doniach model, the enhanced Gamma may be explained by the reduced effective mass in kx-ky plane, where Weyl rings locate. Therefore, this type of layered topological systems may offer a platform to investigate highly anisotropic superconductivity and to understand the extremely large upper critical field in the bulk or in the two-dimensional limit.

cond-mat.supr-con↗