Researcher profile

Yifei Li

Yifei Li contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Artificial Intelligence Machine Learning Computation and Language Computer Vision Graphics physics.plasm-ph Databases math.AP nucl-ex Populations and Evolution

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author

12works

0followers

10topics

4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Claiming links this public author record to a researcher profile and unlocks direct collaboration workflows.

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Inspect adjacent work, topics, institutions and collaborators without jumping out to a separate graph page.

Building this graph slice

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

D3-Gym: Constructing Real-World Verifiable Environments for Data-Driven Discovery

Despite recent progress in language models and agents for scientific data-driven discovery, further advancing their capabilities is held back by the absence of verifiable environments representing real-world scientific tasks. To fill this gap, we introduce D3-Gym, the first automatically constructed dataset with verifiable environments for scientific Data-Driven Discovery. D3-Gym comprises (1) 565 tasks sourced from 239 real scientific repositories across four disciplines where (2) each task is equipped with a natural language instruction, an executable environment with pre-installed dependencies, input dataset and artifact previews, a reference code solution, and an automatically synthesized evaluation script. Rigorous evaluation of the quality of the verification signal in D3-Gym confirms that our evaluation scripts achieve 87.5% agreement with human-annotated gold standards and strong alignment in domain-specific evaluation logic, showing their scientific soundness. Further, training on trajectories sampled from D3-Gym yields consistent and substantial gains across Qwen3 models of varying sizes on ScienceAgentBench, boosting Qwen3-32B by 7.8 absolute points and substantially shrinking the gap with strong proprietary models. All D3-Gym artifacts (environments, creation workflow, trajectories, and models) can be found at https://github.com/OSU-NLP-Group/D3-Gym.

preprint2026arXiv

DepFlow: Disentangled Speech Generation to Mitigate Semantic Bias in Depression Detection

Speech is a scalable and non-invasive biomarker for early mental health screening. However, widely used depression datasets like DAIC-WOZ exhibit strong coupling between linguistic sentiment and diagnostic labels, encouraging models to learn semantic shortcuts. As a result, model robustness may be compromised in real-world scenarios, such as Camouflaged Depression, where individuals maintain socially positive or neutral language despite underlying depressive states. To mitigate this semantic bias, we propose DepFlow, a three-stage depression-conditioned text-to-speech framework. First, a Depression Acoustic Encoder learns speaker- and content-invariant depression embeddings through adversarial training, achieving effective disentanglement while preserving depression discriminability (ROC-AUC: 0.693). Second, a flow-matching TTS model with FiLM modulation injects these embeddings into synthesis, enabling control over depressive severity while preserving content and speaker identity. Third, a prototype-based severity mapping mechanism provides smooth and interpretable manipulation across the depression continuum. Using DepFlow, we construct a Camouflage Depression-oriented Augmentation (CDoA) dataset that pairs depressed acoustic patterns with positive/neutral content from a sentiment-stratified text bank, creating acoustic-semantic mismatches underrepresented in natural data. Evaluated across three depression detection architectures, CDoA improves macro-F1 by 9%, 12%, and 5%, respectively, consistently outperforming conventional augmentation strategies in depression Detection. Beyond enhancing robustness, DepFlow provides a controllable synthesis platform for conversational systems and simulation-based evaluation, where real clinical data remains limited by ethical and coverage constraints.

preprint2026arXiv

Dual-Dimensional Consistency: Balancing Budget and Quality in Adaptive Inference-Time Scaling

Large Language Models (LLMs) have demonstrated remarkable abilities in reasoning. However, maximizing their potential through inference-time scaling faces challenges in trade-off between sampling budget and reasoning quality. Current strategies remain inefficient as they typically treat sampling width and depth as orthogonal objectives, where width consensus methods risk reinforcing hallucinations, while depth pruning mechanisms prematurely truncate complex yet valid reasoning chains. Therefore, we propose Dual-Dimensional Consistency (DDC), a unified framework that bridges path quality with adaptive termination. By coupling Confidence-Weighted Bayesian protocol with a Trend-Aware Stratified Pruning, our method ensures that computational resources are concentrated on high quality reasoning paths, filtering hallucinations while accelerating consensus. Evaluations across five benchmarks demonstrate that this approach reduces token consumption by over 10 times while maintaining or exceeding the accuracy of strong baselines across various LLMs.

preprint2026arXiv

SketchVL: Policy Optimization via Fine-Grained Credit Assignment for Chart Understanding and More

Charts are high-density visual carriers of complex data and medium for information extraction and analysis. Due to the need for precise and complex visual reasoning, automated chart understanding poses a significant challenge to existing Multimodal Large Language Models (MLLMs). Many MLLMs trained with reinforcement learning (RL) face the challenge of credit assignment. Their advantage estimation, typically performed at the trajectory level, cannot distinguish between correct and incorrect reasoning steps within a single generated response. To address this limitation, we introduce SketchVL, a novel MLLM that optimized with FinePO, a new RL algorithm designed for fine-grained credit assignment within each trajectory. SketchVL's methodology involves drawing its intermediate reasoning steps as markers on the image and feeding the annotated image back to itself, creating a robust, multi-step reasoning process. During training, the FinePO algorithm leverages a Fine-grained Process Reward Model (FinePRM) to score each drawing action within a trajectory, thereby precisely assigning credit for each step. This mechanism allows FinePO to more strongly reward correct tokens when a trajectory is globally successful, and more heavily penalize incorrect tokens when the trajectory is globally suboptimal, thus achieving fine-grained reinforcement signals. Experiments show that SketchVL learns to align its step-level behavior with the FinePRM, achieving an average performance gain of 7.23\% over its base model across chart datasets, natural image datasets, and mathematics, providing a promising new direction for training powerful reasoning models.

preprint2026arXiv

ThriftLLM: On Cost-Effective Selection of Large Language Models for Classification Queries

In recent years, large language models (LLMs) have demonstrated remarkable capabilities in comprehending and generating natural language content, attracting widespread attention in both industry and academia. An increasing number of services offer LLMs for various tasks via APIs. Different LLMs demonstrate expertise in different domains of queries (e.g., text classification queries). Meanwhile, LLMs of different scales, complexities, and performance are priced diversely. Driven by this, several researchers are investigating strategies for selecting an ensemble of LLMs, aiming to decrease overall usage costs while enhancing performance. However, to the best of our knowledge, none of the existing works addresses the problem, how to find an LLM ensemble subject to a cost budget, which maximizes the ensemble performance with guarantees. In this paper, we formalize the performance of an ensemble of models (LLMs) using the notion of correctness probability, which we formally define. We develop an approach for aggregating responses from multiple LLMs to enhance ensemble performance. Building on this, we formulate the Optimal Ensemble Selection problem of selecting a set of LLMs subject to a cost budget that maximizes the overall correctness probability. We show that the correctness probability function is non-decreasing and non-submodular and provide evidence that the Optimal Ensemble Selection problem is likely to be NP-hard. By leveraging a submodular function that upper bounds correctness probability, we develop an algorithm called ThriftLLM and prove that it achieves an instance-dependent approximation guarantee with high probability. Our framework functions as a data processing system that selects appropriate LLM operators to deliver high-quality results under budget constraints.

preprint2022arXiv

DiffCloth: Differentiable Cloth Simulation with Dry Frictional Contact

Cloth simulation has wide applications in computer animation, garment design, and robot-assisted dressing. This work presents a differentiable cloth simulator whose additional gradient information facilitates cloth-related applications. Our differentiable simulator extends a state-of-the-art cloth simulator based on Projective Dynamics (PD) and with dry frictional contact. We draw inspiration from previous work to propose a fast and novel method for deriving gradients in PD-based cloth simulation with dry frictional contact. Furthermore, we conduct a comprehensive analysis and evaluation of the usefulness of gradients in contact-rich cloth simulation. Finally, we demonstrate the efficacy of our simulator in a number of downstream applications, including system identification, trajectory optimization for assisted dressing, closed-loop control, inverse design, and real-to-sim transfer. We observe a substantial speedup obtained from using our gradient information in solving most of these applications.

preprint2022arXiv

Femtosecond pumping of nuclear isomeric states by the Coulomb collision of ions with quivering electrons

Efficient production of nuclear isomers is critical for pioneering applications, like nuclear clocks, nuclear batteries, clean nuclear energy, and nuclear γ-ray lasers. However, due to small production cross sections and quick decays, it is extremely difficult to acquire a significant amount of isomers with short lifetimes via traditional accelerators or reactors because of low beam intensity. Here, for the first time, we experimentally present femtosecond pumping of nuclear isomeric states by the Coulomb excitation of ions with the quivering electrons induced by laser fields. Nuclei populated on the third excited state of 83Kr are generated with a peak efficiency of 2.34*10^15 particles=s from a tabletop hundred-TW laser system. It can be explained by the Coulomb excitation of ions with the quivering electrons during the interaction between laser pulses and clusters at nearly solid densities. This efficient and universal production method can be widely used for pumping isotopes with excited state lifetimes down to picoseconds, and could be a benefit for fields like nuclear transition mechanisms and nuclear γ-ray lasers.

preprint2022arXiv

Input-Tuning: Adapting Unfamiliar Inputs to Frozen Pretrained Models

Recently the prompt-tuning paradigm has attracted significant attention. By only tuning continuous prompts with a frozen pre-trained language model (PLM), prompt-tuning takes a step towards deploying a shared frozen PLM to serve numerous downstream tasks. Although prompt-tuning shows good performance on certain natural language understanding (NLU) tasks, its effectiveness on natural language generation (NLG) tasks is still under-explored. In this paper, we argue that one of the factors hindering the development of prompt-tuning on NLG tasks is the unfamiliar inputs (i.e., inputs are linguistically different from the pretraining corpus). For example, our preliminary exploration reveals a large performance gap between prompt-tuning and fine-tuning when unfamiliar inputs occur frequently in NLG tasks. This motivates us to propose input-tuning, which fine-tunes both the continuous prompts and the input representations, leading to a more effective way to adapt unfamiliar inputs to frozen PLMs. Our proposed input-tuning is conceptually simple and empirically powerful. Experimental results on seven NLG tasks demonstrate that input-tuning is significantly and consistently better than prompt-tuning. Furthermore, on three of these tasks, input-tuning can achieve a comparable or even better performance than fine-tuning.

preprint2022arXiv

Interpreting how nonlinear diffusion affects the fate of bistable populations using a discrete modelling framework

Understanding whether a population will survive and flourish or become extinct is a central question in population biology. One way of exploring this question is to study population dynamics using reaction-diffusion equations, where migration is usually represented as a linear diffusion term, and birth-death is represented with a bistable source term. While linear diffusion is most commonly employed to study migration, there are several limitations of this approach, such as the inability of linear diffusion-based models to predict a well-defined population front. One way to overcome this is to generalise the constant diffusivity, $D$, to a nonlinear diffusivity function $D(C)$, where $C>0$ is the density. While it has been formally established that the choice of $D(C)$ affects long-term survival or extinction of a bistable population, working solely in a classical continuum framework makes it difficult to understand precisely how the choice of $D(C)$ affects survival or extinction. Here, we address this question by working with a simple discrete simulation model that is easy to interpret. The continuum limit of the discrete model is a nonlinear reaction-diffusion equation, where the flux involves a nonlinear diffusion term and the source term is given by the strong Allee effect bistable model. We study population extinction/survival using this very intuitive discrete framework together with numerical solutions of the reaction-diffusion continuum limit equation. This approach provides clear insight into how the choice of $D(C)$ either encourages or suppresses population extinction relative to the classical linear diffusion model.

preprint2022arXiv

JoinABLe: Learning Bottom-up Assembly of Parametric CAD Joints

Physical products are often complex assemblies combining a multitude of 3D parts modeled in computer-aided design (CAD) software. CAD designers build up these assemblies by aligning individual parts to one another using constraints called joints. In this paper we introduce JoinABLe, a learning-based method that assembles parts together to form joints. JoinABLe uses the weak supervision available in standard parametric CAD files without the help of object class labels or human guidance. Our results show that by making network predictions over a graph representation of solid models we can outperform multiple baseline methods with an accuracy (79.53%) that approaches human performance (80%). Finally, to support future research we release the Fusion 360 Gallery assembly dataset, containing assemblies with rich information on joints, contact surfaces, holes, and the underlying assembly graph structure.

preprint2022arXiv

Laser plasma accelerated ultra-intense electron beam for efficiently exciting nuclear isomers

Utilizing laser plasma wakefield to accelerate ultra-high charge electron beam is critical for many pioneering applications, for example to efficiently produce nuclear isomers with short lifetimes which may be widely used. However, because of the beam loading effect, electron charge in a single plasma bubble is limited in level of hundreds picocoulomb. Here, we experimentally present that a hundred kilo-ampere, twenty nanocoulomb, tens of MeV collimated electron beam is produced from a chain of wakefield acceleration, via a tightly focused intense laser pulse transversely matched in dense plasma. This ultra-intense electron beam ascribes to a novel efficient injection that the nitrogen atom inner shell electrons are ionized and continuously injected into multiple plasma bubbles. This intense electron beam has been utilized to exciting nuclear isomers with an ultra-high peak efficiency of $1.76\times10^{15}$ particles/s via photonuclear reactions. This efficient production method of isomers can be widely used for pumping isotopes with excited state lifetimes down to picosecond, which is benefit for deep understanding nuclear transition mechanisms and stimulating gamma-ray lasers.

preprint2020arXiv

Travelling wave solutions in a negative nonlinear diffusion-reaction model

We use a geometric approach to prove the existence of smooth travelling wave solutions of a nonlinear diffusion-reaction equation with logistic kinetics and a convex nonlinear diffusivity function which changes sign twice in our domain of interest. We determine the minimum wave speed, c*, and investigate its relation to the spectral stability of the travelling wave solutions.

Institution

Affiliation not imported yet

This author record came from a source that does not expose affiliation metadata. Once the author claims the profile or we enrich the record from another provider, this section will link to the concrete institution.

Topic footprint