Monte Carlo Tree Search (MCTS) has been widely used for automated reasoning data exploration, but current supervision extraction methods remain inefficient. Standard approaches retain only the single highest-reward trajectory, discarding the comparative signals present in the many explored paths. Here we introduce Contrastive Reasoning Path Synthesis (CRPS), a framework that transforms supervision extraction from a filtering process into a synthesis procedure. CRPS uses a structured reflective process to analyze the differences between high- and low-quality search trajectories, extracting explicit information about strategic pivots and local failure modes. These insights guide the synthesis of reasoning chains that incorporate success patterns while avoiding identified pitfalls. We show empirically that models fine-tuned on just 60K CRPS-synthesized examples match or exceed the performance of baselines trained on 590K examples derived from standard rejection sampling, a 20× reduction in dataset size. Furthermore, CRPS improves generalization on out-of-domain benchmarks, demonstrating that learning from the contrast between success and failure produces more transferable reasoning capabilities than learning from success alone.

Supervised Fine-Tuning (SFT) of large language models often suffers from task interference and catastrophic forgetting. Recent approaches alleviate this issue by isolating task-critical parameters during training. However, these methods represent a static solution to a dynamic problem, assuming that parameter importance remains fixed once identified. In this work, we empirically demonstrate that parameter importance exhibits temporal drift over the course of training. To address this, we propose Evolving Parameter Isolation (EPI), a fine-tuning framework that adapts isolation decisions based on online estimates of parameter importance. Instead of freezing a fixed subset of parameters, EPI periodically updates isolation masks using gradient-based signals, enabling the model to protect emerging task-critical parameters while releasing outdated ones to recover plasticity. Experiments on diverse multi-task benchmarks demonstrate that EPI consistently reduces interference and forgetting compared to static isolation and standard fine-tuning, while improving overall generalization. Our analysis highlights the necessity of synchronizing isolation mechanisms with the evolving dynamics of learning diverse abilities.

Supervised Fine-Tuning (SFT) is the standard approach for adapting large language models (LLMs) to downstream tasks. However, we observe a persistent failure mode: even after convergence, models often fail to correctly reproduce a subset of their own supervised training data. We refer to this behavior as the Incomplete Learning Phenomenon (ILP). This paper presents the first systematic study of ILP in LLM fine-tuning. We formalize ILP as post-training failure to internalize supervised instances and demonstrate its prevalence across multiple model families, domains, and datasets. Through controlled analyses, we identify five recurrent sources of incomplete learning: (1) missing prerequisite knowledge in the pre-trained model, (2) conflicts between SFT supervision and pre-training knowledge, (3) internal inconsistencies within SFT data, (4) left-side forgetting during sequential fine-tuning, and (5) insufficient optimization for rare or complex patterns. We introduce a diagnostic-first framework that maps unlearned samples to these causes using observable training and inference signals, and study several targeted mitigation strategies as causal interventions. Experiments on Qwen, LLaMA, and OLMo2 show that incomplete learning is widespread and heterogeneous, and that improvements in aggregate metrics can mask persistent unlearned subsets. The findings highlight the need for fine-grained diagnosis of what supervised fine-tuning fails to learn, and why.

Reward Models (RMs) are key components for evaluating and guiding language model outputs. However, traditional scalar RMs often struggle with incorporating contextual and background information during inference, leading to incomplete evaluations. Generative RMs (GRMs) attempt to address these limitations by generating intermediate reasoning steps. Yet, their uncontrolled black-box nature and inefficiency due to sequential decoding hinder their industrial deployment. Industrial scenarios, such as search and recommendation systems, often involve single-domain tasks requiring evaluation along specific dimensions. In such contexts, diagnosing “bad cases” necessitates structured feedback to identify and optimize dimension-specific issues.In this paper, we propose the Structural Reward Model (SRM), a modular and interpretable framework integrating side-branch models as auxiliary feature generators. By introducing fine-grained dimensions, SRMs enable interpretable and efficient evaluation, facilitating targeted diagnostics and optimization. This structured approach ensures adaptability and scalability for industrial applications.Through comprehensive experiments, we demonstrate that SRMs outperform scalar RMs and GRMs in robustness and alignment with human preferences. The modular design further supports efficient optimization for practical scenarios, allowing SRM to provide a practical reward modeling solution for industry.

Current text mining models are trained with 0-1 hard label that indicates whether an instance belongs to a class, ignoring rich information of the relevance degree. Soft label, which involved each label of varying degrees than the hard label, is considered more suitable for describing instances. The process of generating soft labels from hard labels is defined as label smoothing (LS). Classical LS methods focus on universal data mining tasks so that they ignore the valuable text features in text mining tasks. This paper presents a novel keyword-based LS method to automatically generate soft labels from hard labels via exploiting the relevance between labels and text instances. Generated soft labels are then incorporated into existing models as auxiliary targets during the training stage, capable of improving models without adding any extra parameters. Results of extensive experiments on text classification and large-scale text retrieval datasets demonstrate that soft labels generated by our method contain rich knowledge of text features, improving the performance of corresponding models under both balanced and unbalanced settings.

The embedding-based large-scale query-document retrieval problem is a hot topic in the information retrieval (IR) field. Considering that pre-trained language models like BERT have achieved great success in a wide variety of NLP tasks, we present a QuadrupletBERT model for effective and efficient retrieval in this paper. Unlike most existing BERT-style retrieval models, which only focus on the ranking phase in retrieval systems, our model makes considerable improvements to the retrieval phase and leverages the distances between simple negative and hard negative instances to obtaining better embeddings. Experimental results demonstrate that our QuadrupletBERT achieves state-of-the-art results in embedding-based large-scale retrieval tasks.

Current embedding-based large-scale retrieval models are trained with 0-1 hard label that indicates whether a query is relevant to a document, ignoring rich information of the relevance degree. This paper proposes to improve embedding-based retrieval from the perspective of better characterizing the query-document relevance degree by introducing label enhancement (LE) for the first time. To generate label distribution in the retrieval scenario, we design a novel and effective supervised LE method that incorporates prior knowledge from dynamic term weighting methods into contextual embeddings. Our method significantly outperforms four competitive existing retrieval models and its counterparts equipped with two alternative LE techniques by training models with the generated label distribution as auxiliary supervision information. The superiority can be easily observed on English and Chinese large-scale retrieval tasks under both standard and cold-start settings.