arXiv subjects

Jennifer Yuntong Zhang

Publications and source records attributed to Jennifer Yuntong Zhang.

6 recordsLinked to original sources

arXiv2026

WS-GRPO: Weakly-Supervised Group-Relative Policy Optimization for Rollout-Efficient Reasoning

Group Relative Policy Optimization (GRPO) is effective for training language models on complex reasoning. However, since the objective is defined relative to a group of sampled trajectories, extended deliberation can create more chances to realize relative gains, leading to inefficient reasoning and overthinking, and complicating the trade-off between correctness and rollout efficiency. Controlling this behavior is difficult in practice, considering (i) Length penalties are hard to calibrate because longer rollouts may reflect harder problems that require longer reasoning, penalizing tokens risks truncating useful reasoning along with redundant continuation; and (ii) supervision that directly indicates when to continue or stop is typically unavailable beyond final answer correctness. We propose Weakly Supervised GRPO (WS-GRPO), which improves rollout efficiency by converting terminal rewards into correctness-aware guidance over partial trajectories. Unlike global length penalties that are hard to calibrate, WS-GRPO trains a preference model from outcome-only correctness to produce prefix-level signals that indicate when additional continuation is beneficial. Thus, WS-GRPO supplies outcome-derived continue/stop guidance, reducing redundant deliberation while maintaining accuracy. We provide theoretical results and empirically show on reasoning benchmarks that WS-GRPO substantially reduces rollout length while remaining competitive with GRPO baselines.

cs.LG↗

arXiv2026

Cell-JEPA: Latent Representation Learning for Single-Cell Transcriptomics

Single-cell foundation models learn by reconstructing masked gene expression, implicitly treating technical noise as signal. With dropout rates exceeding 90%, reconstruction objectives encourage models to encode measurement artifacts rather than stable cellular programs. We introduce Cell-JEPA, a joint-embedding predictive architecture that shifts learning from reconstructing sparse counts to predicting in latent space. The key insight is that cell identity is redundantly encoded across genes. We show predicting cell-level embeddings from partial observations forces the model to learn dropout-robust features. On cell-type clustering, Cell-JEPA achieves 0.72 AvgBIO in zero-shot transfer versus 0.53 for scGPT, a 36% relative improvement. On perturbation prediction within a single cell line, Cell-JEPA improves absolute-state reconstruction but not effect-size estimation, suggesting that representation learning and perturbation modeling address complementary aspects of cellular prediction.

cs.CE↗

arXiv2026

SceneAlign: Aligning Multimodal Reasoning to Scene Graphs in Complex Visual Scenes

Multimodal large language models often struggle with faithful reasoning in complex visual scenes, where intricate entities and relations require precise visual grounding at each step. This reasoning unfaithfulness frequently manifests as hallucinated entities, mis-grounded relations, skipped steps, and over-specified reasoning. Existing preference-based approaches, typically relying on textual perturbations or answer-conditioned rationales, fail to address this challenge as they allow models to exploit language priors to bypass visual grounding. To address this, we propose SceneAlign, a framework that leverages scene graphs as structured visual information to perform controllable structural interventions. By identifying reasoning-critical nodes and perturbing them through four targeted strategies that mimic typical grounding failures, SceneAlign constructs hard negative rationales that remain linguistically plausible but are grounded in inaccurate visual facts. These contrastive pairs are used in Direct Preference Optimization to steer models toward fine-grained, structure-faithful reasoning. Across seven visual reasoning benchmarks, SceneAlign consistently improves answer accuracy and reasoning faithfulness, highlighting the effectiveness of grounding-aware alignment for multimodal reasoning.

cs.CV↗

arXiv2025

Realistic CDSS Drug Dosing with End-to-end Recurrent Q-learning for Dual Vasopressor Control

Reinforcement learning (RL) applications in Clinical Decision Support Systems (CDSS) frequently encounter skepticism because models may recommend inoperable dosing decisions. We propose an end-to-end offline RL framework for dual vasopressor administration in Intensive Care Units (ICUs) that directly addresses this challenge through principled action space design. Our method integrates discrete, continuous, and directional dosing strategies with conservative Q-learning and incorporates a novel recurrent modeling using a replay buffer to capture temporal dependencies in ICU time-series data. Our comparative analysis of norepinephrine dosing strategies across different action space formulations reveals that the designed action spaces improve interpretability and facilitate clinical adoption while preserving efficacy. Empirical results on eICU and MIMIC demonstrate that action space design profoundly influences learned behavioral policies. Compared with baselines, the proposed methods achieve more than 3x expected reward improvements, while aligning with established clinical protocols.

cs.LG↗

arXiv2025

In-Context Algorithm Emulation in Fixed-Weight Transformers

We prove that a minimal Transformer with frozen weights emulates a broad class of algorithms by in-context prompting. We formalize two modes of in-context algorithm emulation. In the task-specific mode, for any continuous function $f: \mathbb{R} \to \mathbb{R}$, we show the existence of a single-head softmax attention layer whose forward pass reproduces functions of the form $f(w^\top x - y)$ to arbitrary precision. This general template subsumes many popular machine learning algorithms (e.g., gradient descent, linear regression, ridge regression). In the prompt-programmable mode, we prove universality: a single fixed-weight two-layer softmax attention module emulates all algorithms from the task-specific class (i.e., each implementable by a single softmax attention) via only prompting. Our key idea is to construct prompts that encode an algorithm's parameters into token representations, creating sharp dot-product gaps that force the softmax attention to follow the intended computation. This construction requires no feed-forward layers and no parameter updates. All adaptation happens through the prompt alone. Numerical results corroborate our theory. These findings forge a direct link between in-context learning and algorithmic emulation, and offer a simple mechanism for large Transformers to serve as prompt-programmable libraries of algorithms. They illuminate how GPT-style foundation models may swap algorithms via prompts alone, and establish a form of algorithmic universality in modern Transformer models.

cs.LG↗

arXiv2025

Are Hallucinations Bad Estimations?

We formalize hallucinations in generative models as failures to link an estimate to any plausible cause. Under this interpretation, we show that even loss-minimizing optimal estimators still hallucinate. We confirm this with a general high probability lower bound on hallucinate rate for generic data distributions. This reframes hallucination as structural misalignment between loss minimization and human-acceptable outputs, and hence estimation errors induced by miscalibration. Experiments on coin aggregation, open-ended QA, and text-to-image support our theory.

cs.LG↗