Zhikun Xu

ri @AIatMeta, pursuing CS PhD @SCAI_ASU | Prev: Applied Math (B.S. & M.S.) @FudanUni, Research Internship @awscloud @AlibabaGroup @AMD | Opinions are my own.

[Zhikun Xu @JerrryKun]

@xwang_lk

[Zhikun Xu @JerrryKun]

@hendav136 🫡 (pretraining for what?

[Zhikun Xu @JerrryKun]

@wenhaocha1 Indeed. Tokens are becoming electricity. Acting is more expensive.

[Zhikun Xu @JerrryKun]

https://t.co/PeJmwq87ec

[Zhikun Xu @JerrryKun]

Treating reasoning and acting as two tools for one job folds many debates (long context vs RAG, think vs act, etc.) into a single allocation question. And framing the internalization–externalization boundary as the next design question feels exactly right. Really inspiring read!

[Hongru Wang @HongruWang007]

Two students take the same exam. Both score 100 — one solved it himself, the other Googled every answer.

A semester later, the gap is huge.

That's the problem with today's AI agents. I write a detailed blog to share my recent thoughts on this, mainly based on Theory of Agents. I promise this is definitely worth 30 minutes of your time.

Blog: https://t.co/VCFC7RnbU6
Project: https://t.co/WFLEYhOaCl

[Zhikun Xu @JerrryKun]

That last part resonates with what motivated our ongoing project, ReuseRL: if a capability only internalizes when it is compressible, the real question becomes what gets internalized. We go looking for the atoms: the small set of reusable skills a model can absorb and build on.

[Zhikun Xu @JerrryKun]

@Swarooprm7 We may need to define the "machine creativity/novelty" first. What about searching for some counterexamples in formal math? Concepts are created to simplify things, facilitate abstraction, and improve human understanding (efficiency, etc.). Why do AI models need novel concepts?

[Zhikun Xu @JerrryKun]

@wzenus @YejinChoinka @jiajunwu_cs @ManlingLi_ @LINJIEFUN @chi_gui_1 @DeimosGN @qineng_wang @James_KKW @shiqi_chen17 @zhengyuan_yang Well-deserved👍

[Zhikun Xu @JerrryKun]

✨Check out the paper to learn more! This paper was done by our intern @Zijun0916 last summer. A big thank you to my labmate @XiaoYe1170354 and our advisor @BenZhou96 for the support and guidance!

[Zhikun Xu @JerrryKun]

LLMs memorize massive amounts of text, but can they actually apply this knowledge conceptually? 🤔

Our #ICLR 26' paper from the ARC Lab probes this in math reasoning! "CORE: Concept-Oriented Reinforcement for Bridging the Definition-Application Gap..." 🔗 https://t.co/GscQjqjTMd

[Zhikun Xu @JerrryKun]

📈 The results: Consistent gains over vanilla baselines, including up to +9.3% on in-domain Textbook problems and +9.6% on out-of-domain TheoremQA. We also did ablation experiments to show the results are consistent with different models and across different benchmarks.

[Zhikun Xu @JerrryKun]

@HBX_hbx @QuYuxiao Besides QuestA, there are also many other related works using a similar idea from last year: BREAD(https://t.co/O7wOGJnYPD), Scaf-GRPO(https://t.co/oug6ytphJ3), and CORE(https://t.co/1ygNSBsQC1). "Guided prefix" could be partial oracle solutions, problem-related concepts, etc.

[Zhikun Xu @JerrryKun]

@denny_zhou Fake agi: Gemini🤖
Real agi: daughter🧒

[Zhikun Xu @JerrryKun]

Highly resonate about this. Conceptual reasoning should be very important for LLM reasoning.

[Denny Zhou @denny_zhou]

To the questions of “why not both?”: my dream is for LLMs to make conceptual discoveries, like Galois with group theory or Einstein with general relativity. I don’t believe breakthroughs like these would come from A* search or its more advanced version MCTS.

[Zhikun Xu @JerrryKun]

@ShashwatGoel7 👀 https://t.co/0ZaHOGxxEx

[Zhikun Xu @JerrryKun]

First time at #NeurIPS2025! I’ll be in San Diego from Dec 1–6 and would love to make new friends, grab some tea🍵, and discuss LLM reasoning (math & cognition-inspired), RL, and more! Feel free to DM!

[Zhikun Xu @JerrryKun]

@zjasper @hyperbolic_labs https://t.co/H5u5ZB51HL 👀how about testing it with some counterexamples?

[Zhikun Xu @JerrryKun]

@taiwei_shi NeurIPS+1! Hope to talk with u🙋