Research
Turning frontline deployments into reproducible research.
Retrieval-augmented generation, LLM long-term memory, and knowledge graphs: 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
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 It reconnects two lineages that do not cite each other: the 1999–2017 interruptibility literature, which formalized interruption cost rigorously but had no capable actor, and the 2024–2026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: until intervention quality is measurable, a reinforcement-learning reward for proactivity cannot be defined. Proposes the design of a benchmark for open-world intervention timing with an explicit cost term.
Registers, Not Plans: What Lives in a Language Model's Workspace That Isn't on Its Tongue
Tao An
An independent replication of Anthropic's global-workspace claim on open-weight models. Filtering lens readouts by the model's own next-token distribution splits the workspace in two: causally steerable context registers (the conversation's language, a corrected typo) survive, while content plans (rhyme, arithmetic) fall to the permutation floor. Editing a register also rewrites the model's representation of the question it was asked, stably across a 1.7B–14B ladder and a second architecture.
CGEP: Toward Detecting and Attributing GEO Poisoning in Chinese AI Search
Tao An
Defines GEO-poisoning detection and attribution for Chinese generative search (DeepSeek, Doubao, Kimi): a five-technique taxonomy of coordinated inauthentic manipulation, a task reframing from attack-success to detection → classification → account-cluster attribution, and a legally-constructed synthetic benchmark. A provenance pilot shows detection and attribution need different substrates: content features detect that manipulation happened (F1 0.93) but only an account-interaction graph attributes it to a seller cluster (0.96), and a confidence-gated fusion covers the taxonomy where a learned GNN and a zero-shot LLM both fail.
The Preference Centroid: Consensus Density Governs Output Dispersion in Aligned LLMs
Tao An, Shuai Feng
Sampling an aligned LLM repeatedly and embedding the completions, output dispersion is governed by the consensus density of the task: near-zero on factual prompts, wide on open-ended ones (Spearman ρ = 0.85), replicating on a second model and predicted by held-out judges that score only the prompt (ρ = −0.91). A base-vs-instruct comparison shows alignment amplifies a gradient the pretrained base already carries.
It's Fidelity, Not 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.
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.
Cognitive Workspace: Active Memory Management for LLMs
Tao An
Proposes Cognitive Workspace, a paradigm transcending traditional RAG by emulating human cognition: active memory management, hierarchical cognitive buffers, and task-driven context optimization. Achieves a 58.6% memory-reuse rate (vs. 0% for RAG) with a 17–18% net efficiency gain.
A Graph-Neural-Network Method for Data-Information Recommendation
Tao An
Chinese invention patent, under examination. GNN-based recommendation over heterogeneous data–information graphs.
Academic service
Ethics Reviewer
NeurIPS 2026· 2026Ethics 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