Researcher profile

Shan Li

Shan Li contributes to research discovery and scholarly infrastructure.

ResearcherAffiliation not importedOpen to collaborate

Trust snapshot

Quick read

Trust 21 - EmergingVerification L1Unclaimed author
9works
0followers
12topics
4close collaborators

Actions

Decide how to stay connected

Follow researcher0

Identity and collaboration

How to connect with this researcher

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

Log in to claim

Direct collaboration

Open a focused conversation when the fit is right

Claim this author entity first to unlock direct invitations.

Research graph

See the researcher in context

Open full explorer

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.

Published work

9 published item(s)

preprint2026arXiv

Dynamic Adversarial Fine-Tuning Reorganizes Refusal Geometry

Safety-aligned language models must refuse harmful requests without collapsing into broad over-refusal, yet it remains unclear how dynamic adversarial fine-tuning changes the internal carriers of refusal. We study one 7B backbone under supervised fine-tuning (SFT) and under Robust Refusal Dynamic Defense (R2D2), a HarmBench-style adversarial fine-tuning procedure that repeatedly refreshes harmful training cases with current jailbreak attacks. Our protocol aligns fixed-source HarmBench, StrongREJECT, and XSTest with a five-anchor refusal-geometry suite, causal interventions, and a sparse adaptive stress test. R2D2 drives fixed-source HarmBench attack success to zero at early checkpoints, but that regime coincides with maximal XSTest refusal and complete failure on a benign-utility audit. Later checkpoints partially recover benign utility while partially reopening attack success. Sparse adaptive attacks sharpen the same frontier: step~50 remains closed under both adaptive GCG and AutoDAN, whereas adaptive GCG ASR rises to 0.415 at step~250 and 0.613 at step~500. Geometrically, R2D2 preserves a late-layer admissible carrier through step~100 and relocates the best admissible carrier to an early layer by step~250; SFT relocates earlier while remaining less robust. Effective rank remains near 1.24, and SFT exhibits larger principal-angle drift despite worse robustness. Causal interventions show that late-stage R2D2 behavior is controlled by a low-dimensional but utility-coupled carrier. These results support a geometry-reorganization account along a robustness--utility frontier.

preprint2020arXiv

Deep Job Understanding at LinkedIn

As the world's largest professional network, LinkedIn wants to create economic opportunity for everyone in the global workforce. One of its most critical missions is matching jobs with processionals. Improving job targeting accuracy and hire efficiency align with LinkedIn's Member First Motto. To achieve those goals, we need to understand unstructured job postings with noisy information. We applied deep transfer learning to create domain-specific job understanding models. After this, jobs are represented by professional entities, including titles, skills, companies, and assessment questions. To continuously improve LinkedIn's job understanding ability, we designed an expert feedback loop where we integrated job understanding models into LinkedIn's products to collect job posters' feedback. In this demonstration, we present LinkedIn's job posting flow and demonstrate how the integrated deep job understanding work improves job posters' satisfaction and provides significant metric lifts in LinkedIn's job recommendation system.

preprint2020arXiv

Half-Heusler thermoelectric materials: NMR studies

We report $^{59}$Co, $^{93}$Nb, and $^{121}$Sb nuclear magnetic resonance (NMR) measurements combined with density functional theory (DFT) calculations on a series of half-Heusler semiconductors, including NbCoSn, ZrCoSb, TaFeSb and NbFeSb, to better understand their electronic properties and general composition-dependent trends. These materials are of interest as potentially high efficiency thermoelectric materials. Compared to the other materials, we find that ZrCoSb tends to have a relatively large amount of local disorder, apparently antisite defects. This contributes to a small excitation gap corresponding to an impurity band near the band edge. In NbCoSn and TaFeSb, Curie-Weiss-type behavior is revealed, which indicates a small density of interacting paramagnetic defects. Very large paramagnetic chemical shifts are observed associated with a Van Vleck mechanism due to closely spaced $d$ bands splitting between the conduction and valence bands. Meanwhile, DFT methods were generally successful in reproducing the chemical shift trend for these half-Heusler materials, and we identify an enhancement of the larger-magnitude shifts, which we connect to electron interaction effects. The general trend is connected to changes in $d$-electron hybridization across the series.

preprint2020arXiv

Learning to Ask Screening Questions for Job Postings

At LinkedIn, we want to create economic opportunity for everyone in the global workforce. A critical aspect of this goal is matching jobs with qualified applicants. To improve hiring efficiency and reduce the need to manually screening each applicant, we develop a new product where recruiters can ask screening questions online so that they can filter qualified candidates easily. To add screening questions to all $20$M active jobs at LinkedIn, we propose a new task that aims to automatically generate screening questions for a given job posting. To solve the task of generating screening questions, we develop a two-stage deep learning model called Job2Questions, where we apply a deep learning model to detect intent from the text description, and then rank the detected intents by their importance based on other contextual features. Since this is a new product with no historical data, we employ deep transfer learning to train complex models with limited training data. We launched the screening question product and our AI models to LinkedIn users and observed significant impact in the job marketplace. During our online A/B test, we observed $+53.10\%$ screening question suggestion acceptance rate, $+22.17\%$ job coverage, $+190\%$ recruiter-applicant interaction, and $+11$ Net Promoter Score. In sum, the deployed Job2Questions model helps recruiters to find qualified applicants and job seekers to find jobs they are qualified for.

preprint2019arXiv

A Deeper Look at Facial Expression Dataset Bias

Datasets play an important role in the progress of facial expression recognition algorithms, but they may suffer from obvious biases caused by different cultures and collection conditions. To look deeper into this bias, we first conduct comprehensive experiments on dataset recognition and crossdataset generalization tasks, and for the first time explore the intrinsic causes of the dataset discrepancy. The results quantitatively verify that current datasets have a strong buildin bias and corresponding analyses indicate that the conditional probability distributions between source and target datasets are different. However, previous researches are mainly based on shallow features with limited discriminative ability under the assumption that the conditional distribution remains unchanged across domains. To address these issues, we further propose a novel deep Emotion-Conditional Adaption Network (ECAN) to learn domain-invariant and discriminative feature representations, which can match both the marginal and the conditional distributions across domains simultaneously. In addition, the largely ignored expression class distribution bias is also addressed by a learnable re-weighting parameter, so that the training and testing domains can share similar class distribution. Extensive cross-database experiments on both lab-controlled datasets (CK+, JAFFE, MMI and Oulu-CASIA) and real-world databases (AffectNet, FER2013, RAF-DB 2.0 and SFEW 2.0) demonstrate that our ECAN can yield competitive performances across various facial expression transfer tasks and outperform the state-of-theart methods.

preprint2019arXiv

Knowledge-aided Two-dimensional Autofocus for Spotlight SAR Filtered Backprojection Imagery

Filtered backprojection (FBP) algorithm is a popular choice for complicated trajectory SAR image formation processing due to its inherent nonlinear motion compensation capability. However, how to efficiently autofocus the defocused FBP imagery when the motion measurement is not accurate enough is still a challenging problem. In this paper, a new interpretation of the FBP derivation is presented from the Fourier transform point of view. Based on this new viewpoint, the property of the residual 2-D phase error in FBP imagery is analyzed in detail. Then, by incorporating the derived a priori knowledge on the 2-D phase error, an accurate and efficient 2-D autofocus approach is proposed. The new approach performs the parameter estimation in a dimension-reduced parameter subspace by exploiting the a priori analytical structure of the 2-D phase error, therefore possesses much higher accuracy and efficiency than conventional blind methods. Finally, experimental results clearly demonstrate the effectiveness and robustness of the proposed method.

preprint2018arXiv

Deep Facial Expression Recognition: A Survey

With the transition of facial expression recognition (FER) from laboratory-controlled to challenging in-the-wild conditions and the recent success of deep learning techniques in various fields, deep neural networks have increasingly been leveraged to learn discriminative representations for automatic FER. Recent deep FER systems generally focus on two important issues: overfitting caused by a lack of sufficient training data and expression-unrelated variations, such as illumination, head pose and identity bias. In this paper, we provide a comprehensive survey on deep FER, including datasets and algorithms that provide insights into these intrinsic problems. First, we describe the standard pipeline of a deep FER system with the related background knowledge and suggestions of applicable implementations for each stage. We then introduce the available datasets that are widely used in the literature and provide accepted data selection and evaluation principles for these datasets. For the state of the art in deep FER, we review existing novel deep neural networks and related training strategies that are designed for FER based on both static images and dynamic image sequences, and discuss their advantages and limitations. Competitive performances on widely used benchmarks are also summarized in this section. We then extend our survey to additional related issues and application scenarios. Finally, we review the remaining challenges and corresponding opportunities in this field as well as future directions for the design of robust deep FER systems.