Research

Turning frontline deployments into reproducible research.

Human-AI collaboration, LLM memory, metascience, and agent safety: formalizing production practice into reproducible work. Below are the team's public papers, patents, and academic service, each linked to a verifiable source.

Papers & patents

  • EACL 2027 · Main (to appear)LLM Memory2027

    Fidelity Before Structure: Verbatim Chunks Beat Lossy Artifact Extraction in Long-Conversation LLM Memory

    Tao An

    A controlled ablation isolating the stored memory representation inside one fixed retrieve–rerank–reason pipeline: LLM-extracted typed artifacts versus verbatim conversation chunks. Verbatim chunks win by 15.9 points on LoCoMo (43.9% vs. 28.0%) and 22.0 points on LongMemEval-S. Structured memory should augment verbatim text, not replace it. (Formerly titled It's Fidelity, Not Structure.)
  • TMLR 2026 · SurveyHuman–AI2026

    When Should the Agent Speak? A Survey of Intervention Timing for Always-On AI Assistants

    Tao An

    A survey of intervention timing for always-on assistants (smart glasses, MR headsets, ambient copilots), organized around one decision rule: intervene iff the expected benefit of acting exceeds the expected cost of interrupting. It reconnects two lineages that do not cite each other: the 1999–2017 interruptibility literature, which formalized interruption cost rigorously but had no capable actor, and the 2024–2026 proactive-agent wave, which has actors but rediscovers the cost term only in fragments (false-alarm pricing, cognitive load, social violation, compute) that no work unifies. Evaluation is the gating layer: mobile health already learns intervention timing online against a validated proximal outcome, while open-world assistance has no analogous outcome, so a reinforcement-learning objective for proactivity has no shared, validated reward to train against. Proposes the design of a benchmark for open-world intervention timing with an explicit cost term.
  • TMLR 2026Human–AI2026

    Consensus Density Predicts Output Dispersion in Aligned LLMs

    Tao An, Shuai Feng

    Sampling an aligned LLM 25 times per prompt and embedding the completions, output dispersion (mean pairwise cosine distance) is predicted by the consensus density of the task: held-out judges that see only the prompt predict it at Spearman ρ = −0.91. The task-type ordering holds on two unrelated models (ρ = 0.85 and 0.84), the judge relation holds within two professional domains (−0.83 over 42 pharmaceutical prompts, −0.62 over 40 software-engineering prompts), and a preliminary expert panel tracks the judges at ρ = 0.74. A base-vs-instruct comparison shows alignment amplifies a gradient the pretrained base already carries. The link to AI equalizing or amplifying human outcomes is stated separately, as an untested hypothesis.
  • HHAI 2026 · IOS PressHuman–AI2026

    AI as Equalizer or Amplifier? Task Complexity as the Moderating Factor for Human Expertise in Hybrid Intelligence Systems

    Tao An

    Published in the proceedings of the 5th International Conference on Hybrid Human-Artificial Intelligence (HHAI 2026, Brussels), IOS Press Frontiers in AI and Applications vol. 423, pp. 212–220 (open access, CC BY-NC). Drawing on structured field observations since mid-2024, this position paper reconciles the 'equalizer' and 'amplifier' debates: AI narrows novice–expert gaps on routine tasks but amplifies them on complex tasks requiring deep judgment. Domain expertise, not prompt engineering, determines who benefits most.
  • NeurIPS 2026 Workshop · AI4MetaScience · Poster (non-archival)Metascience2026

    Recombination or Discovery? A Retrieval-Grounded Novelty Audit of Machine-Generated Research Papers

    Tao An

    Poster at the NeurIPS 2026 Workshop on AI for Meta-Science (AI4MetaScience; non-archival). A retrieval-grounded, protocol-frozen novelty audit of 166 machine-generated papers (FARS) against 166 topic-matched ICLR 2025 submissions. Each paper is decomposed into contribution claims on four facets (purpose, mechanism, evaluation, domain; 549 machine and 494 human contributions); prior art is retrieved under per-paper submission-date cutoffs; contributions are classified as covered, recombination, or facet-novel under a pre-registered two-judge protocol. Machine contributions are judged facet-novel more often than human ones (56.1% vs. 35.6%), and the gap rests mainly on the purpose facet: requiring an uncovered facet other than purpose shrinks it from 20.5 to 4.1 points (95% CI −0.3 to 8.5). The judge layer cannot be certified by another LLM: adversarial re-auditors return refutation rates from 0% to 100% on the same items depending on auditor model and prompt. A 106-pair gold prior-art audit finds the deployed retrieval surfaces the known prior art for only 25–29% of pairs per arm, so automated novelty rates are exploratory upper bounds. A companion integrity audit of 306 Agents4Science 2025 submissions finds hard fabrication evidence in 0/47 accepted versus 16/197 rejected (one-sided Fisher p = 0.029; 0.072 under conservative coding), an association that reflects both AI-reviewer detection and authors' own disclosure.
  • Patent application · Under examinationAI Safety2026

    A Tool-Call Governance Method and System for Large Language Model Agents

    Tao An

    Chinese invention patent application No. 202611257740.2, filed 19 August 2026 by Beijing FIM Network Technology Co., Ltd.; passed preliminary examination on 10 September 2026; published by CNIPA on 17 November 2026 as CN 123116265 A and entered substantive examination the same day. Not yet granted.

Academic service

  • Ethics Reviewer

    NeurIPS 2026· 2026

    Tao An

    Ethics Review Committee, Conference on Neural Information Processing Systems (NeurIPS 2026), reviewing flagged submissions against the NeurIPS Code of Ethics: data provenance and informed consent, dual-use and misuse risk, human-subjects considerations, and broader societal impact.

Want to go deep on a research direction?

Research collaboration, a technical approach, or just a chat about a paper? Reach out directly.

Contact us