Junhua Chen

Aspiring computer science and mathematics student with wide range of interests.

Notes8

• Quant employment verification: CONFIRMED. Two independent primary sources establish real quant roles plus a supported identity. (1) Roles: LinkedIn-derived profile lists Quantitative Researcher at Jane Street (London, 2025- ) as current, Quantitative Research Intern at Jane Street (London, 2023) and Software Developer Internship at Optiver (Sydney, 2022, risk department, end-to-end SDLC) as past; his own site states 'I also work in quantitative finance'. (2) Employer classification: Jane Street is a global quantitative proprietary-trading firm/market maker and Optiver is a global market maker/quantitative trading firm, so both are genuine quantitative-finance employers, not merely 'quant'-named. (3) Identity: IOI stats page shows Junhua Chen competed for Australia at IOI 2020 (Bronze, 299.09 pts, rank 118/343); his LinkedIn lists Caulfield Grammar School (the IOI report names his school) and University of Cambridge, and his personal site ties the same publications and Cambridge math/CS degree to the quant role. Caveat: the CURRENT Jane Street role and the 2025 start rest on the LinkedIn-index record, which flags it current - not independently re-confirmed from a second live source, so treat 'present' as index-reported. The two internships are clearly documented as actual work.Sep 2025
• Recruited into Jane Street as a Quantitative Researcher in London from September 2025, after two prior stints (2023 research internship, 2022 Optiver risk-department software internship) - the direct intern-to-full-time quant pipeline typical of prop-trading firms, meaning his quant career was deliberately seeded from undergraduate internships rather than a late pivot.Sep 2025
• His identity is corroborated across three independent sources - IOI statistics (school, country, medal), LinkedIn (same school, Cambridge, Jane Street), and his personal site (title, publications, Cambridge math/CS) - so the IOI 2020 Australian competitor and the Jane Street researcher are the same person, not a namesake collision.Sep 2025
• First-author paper 'Sequential Controlled Langevin Diffusions' accepted at ICLR 2025 (co-authors Lorenz Richter, Julius Berner, Denis Blessing, Gerhard Neumann, Anima Anandkumar) - a generative-modelling result produced during his Caltech summer research (2024), showing research output already competitive at top ML venues while still an undergraduate.Jan 2025
• Withdrew from a University of Melbourne Chancellor's-Scholar computer-science degree (2021) to move to a BA in Mathematics at Trinity College, Cambridge (2021-2024), then an MPhil in Computer Science at Cambridge (2024-2025) - a research-mathematics-to-generative-modelling arc running in parallel with the quant internships.Sep 2021
• Won a Bronze medal for Australia at IOI 2020 (absolute score 299.09, rank 118/343; relative rank 65.89%), representing Caulfield Grammar School - the entry point that led to tutoring for the Australian Maths Trust Informatics School of Excellence and his competitive-programming trajectory into Cambridge mathematics.Sep 2020
• Published 'Straight but not so fast: Challenges with Rectified Flows in Protein Design' at the ICML 2025 GenBio workshop with the Cambridge Lio group, plus 'Group Policy Gradient' at the CoRL 2025 workshop - his 2025 research spans diffusion models, protein design and reinforcement learning, the signature of a generative-modelling specialist rather than a generic quant.Jul 2025
• Spent summer 2024 as a Summer Undergraduate Research Student at Caltech (Pasadena), a machine-learning research internship supervised by Dr Julius Berner and Prof Anima Anandkumar - the origin of the ICLR 2025 paper and a rare research-scientist credential alongside his trading roles.Jun 2024
• Co-authored 'Beyond Erdos-Renyi: Generalization in Algorithmic Reasoning on Graphs' at the Second Learning on Graphs Conference (2023) with Tunan Shi among the co-authors - the same Tunan Shi who took Silver for Australia at IOI 2020 beside him, indicating his research collaborations grew out of the olympiad cohort.Nov 2023
• Worked as a software engineer in Optiver's risk department in Sydney (2022), owning the full development cycle from stakeholder requirements to production DevOps - a genuine engineering internship at a market maker, distinct from competition or insight-day exposure.Jun 2022
• Runs the personal site anonymous3141.github.io (GitHub handle anonymous3141, 15 public repos, 13 followers) where he declares generative modelling as his research focus ('Currently, I work in machine learning, with a particular emphasis on generative modelling and its applications') alongside the quant role - a dual research/trading identity he states explicitly.Sep 2025
• Declares research interests in multi-modal generative modelling (diffusion vs autoregressive vs novel approaches), fine-tuning/steering of generative models, and their scaling efficiency - a focused agenda rather than a broad ML dabbler, which pairs with the diffusion-model publications.Sep 2025