arXiv subjects

Xinxin Song

Publications and source records attributed to Xinxin Song.

5 recordsLinked to original sources

arXiv2026

CARE: Contrastive Anchor-based Rubric Evolution for Large Language Model Post-Training

Rubric-based reinforcement learning decomposes open-ended instructions into prompt-specific, flexible rubrics, making it better suited than reinforcement learning with verifiable rewards for post-training LLMs on open-ended tasks. However, static rubrics are inevitably hacked as the policy evolves, and existing dynamic approaches introduce new problems: undirected rubric extraction, unreliable hack detection, and unbounded rubric proliferation. We propose $\textbf{CARE}$ ($\textbf{C}$ontrastive $\textbf{A}$nchor-based $\textbf{R}$ubric $\textbf{E}$volution), which grounds every rubric evolution step in a high-quality anchor response generated by a frontier model conditioned on the prompt and its rubrics. At each training step, CARE contrasts the highest-scoring rollout against the anchor, enabling two complementary mechanisms: an Adaptive branch that reactively repairs reward misspecification; and a Chase branch that proactively converts frontier-level quality gaps into sharper rubrics. Together, the two branches $\textbf{maintain discriminative accuracy in the high-reward region}$---the precise region where reward over-optimization mostly originates. Experiments on WildChecklist-9K with Qwen2.5-7B-Base and Qwen2.5-7B-Instruct show that CARE achieves state-of-the-art performance on Arena-Hard-2.0, InfoBench, and FollowBench, and is the $\textbf{only}$ method whose win rate against GPT-4.1 anchor responses shows sustained improvement throughout 300 training steps; additional results on Llama-3.1-8B-Instruct and Qwen3-8B further indicate that CARE generalizes across model families.

cs.AI↗

arXiv2026

Fine-Tuning a Large Vision-Language Model for Artwork's Scoring and Critique

Assessing artistic creativity is foundational to creativity research and arts education, yet manual scoring (e.g., Torrance Tests of Creative Thinking) is labor-intensive at scale. Prior machine-learning approaches show promise for visual creativity scoring, but many rely mainly on image features and provide limited or no explanatory feedback. We propose a framework for automated creativity assessment of human paintings by fine-tuning the vision-language model Qwen2-VL-7B with multi-task learning. Our dataset contains 1000 human-created paintings scored on a 1-100 scale and paired with a short human-written description (content or artist explanation). Two expert raters evaluated each work using a five-dimension rubric (originality, color, texture, composition, content) and provided written critiques; we use an 80/20 train-test split. We add a lightweight regression head on the visual encoder output so the model can predict a numerical score and generate rubric-aligned feedback in a single forward pass. By embedding the structured rubric and the artwork description in the system prompt, we constrain the generated text to match the quantitative prediction. Experiments show strong accuracy, achieving Pearson r > 0.97 and MAE about 3.95 on the 100-point scale. Qualitative evaluation indicates the generated feedback is semantically close to expert critiques (average SBERT cosine similarity = 0.798). The proposed approach bridges computer vision and art assessment and offers a scalable tool for creativity research and classroom feedback.

cs.CV↗

arXiv2025

A Unified Data-Driven Framework for Efficient Scientific Discovery

Scientific discovery drives progress across disciplines, from fundamental physics to industrial applications. However, identifying physical laws automatically from gathered datasets requires identifying the structure and parameters of the formula underlying the data, which involves navigating a vast search space and consuming substantial computational resources. To address these issues, we build on the Buckingham $\Pi$ theorem and Taylor's theorem to create a unified representation of diverse formulas, which introduces latent variables to form a two-stage structure. To minimize the search space, we initially focus on determining the structure of the latent formula, including the relevant contributing inputs, the count of latent variables, and their interconnections. Following this, the process of parameter identification is expedited by enforcing dimensional constraints for physical relevance, favoring simplicity in the formulas, and employing strategic optimization techniques. Any overly complex outcomes are refined using symbolic regression for a compact form. These general strategic techniques drastically reduce search iterations from hundreds of millions to just tens, significantly enhancing the efficiency of data-driven formula discovery. We performed comprehensive validation to demonstrate FIND's effectiveness in discovering physical laws, dimensionless numbers, partial differential equations, and uniform critical system parameters across various fields, including astronomy, physics, chemistry, and electronics. The excellent performances across 11 distinct datasets position FIND as a powerful and versatile tool for advancing data-driven scientific discovery in multiple domains.

cs.CE↗

arXiv2025

Causally-informed Deep Learning towards Explainable and Generalizable Outcomes Prediction in Critical Care

Recent advances in deep learning (DL) have prompted the development of high-performing early warning score (EWS) systems, predicting clinical deteriorations such as acute kidney injury, acute myocardial infarction, or circulatory failure. DL models have proven to be powerful tools for various tasks but come with the cost of lacking interpretability and limited generalizability, hindering their clinical applications. To develop a practical EWS system applicable to various outcomes, we propose causally-informed explainable early prediction model, which leverages causal discovery to identify the underlying causal relationships of prediction and thus owns two unique advantages: demonstrating the explicit interpretation of the prediction while exhibiting decent performance when applied to unfamiliar environments. Benefiting from these features, our approach achieves superior accuracy for 6 different critical deteriorations and achieves better generalizability across different patient groups, compared to various baseline algorithms. Besides, we provide explicit causal pathways to serve as references for assistant clinical diagnosis and potential interventions. The proposed approach enhances the practical application of deep learning in various medical scenarios.

cs.LG↗

arXiv2024

Using a CNN Model to Assess Paintings' Creativity

Assessing artistic creativity has long challenged researchers, with traditional methods proving time-consuming. Recent studies have applied machine learning to evaluate creativity in drawings, but not paintings. Our research addresses this gap by developing a CNN model to automatically assess the creativity of human paintings. Using a dataset of six hundred paintings by professionals and children, our model achieved 90% accuracy and faster evaluation times than human raters. This approach demonstrates the potential of machine learning in advancing artistic creativity assessment, offering a more efficient alternative to traditional methods.

cs.CV↗