Xin Jin

Previous RA at @Westlake_Uni, Master at CTBU.

[Xin Jin @Xander_K1ng]

🎉Glad to announce our survey on Mixup are now released on arXiv🌠
-Nearly 220 papers with lots of training paradigms, data modalities
-Summarize papers based on improvement
-Reformulated mixup as the framework
-Appeal designing a mixup unified framework to reduce the gap
#Mixup

[Huan Wang @ICML26 @huanwangx]

Heading to #ICML2026 in Seoul 🇰🇷with 3 students. See you soon my friends chatting efficient AI chewing kbbq 😃

Check the collection of our 4 papers here: https://t.co/OAc1xNWqg2

[Xin Jin @Xander_K1ng]

@Shiwei_Liu66 @LupinLSY The SGG (https://t.co/HQsAp029Pj) also employs a Layer-wise LR adaptive technique. This looks like a useful technique for LLMs & MLLMs beyond vanilla AdamW or Muon.

[Zhuoli Ouyang @Collapsar0000]

Thrilled to share our #ICML2026 work! We rethink optimizer design by leveraging the row-block diagonal dominance of the Transformer Hessian. We show that simple row-normalization (RMNP) is equivalent to Muon's orthogonalization.

Full story in the thread! 👇

[Shenyang Deng ✈️ ICML2026 @DengShenyang24]

1/n Please stop by👋. This is not just another ICML 2026 optimizer paper. We have rich intuition to share on why simple preconditioners like orthogonalization and row-normalization specifically benefit NNs optimization. Quick overview below 🧵

[Victor.Kai Wang @VictorKaiWang1]

Glad to see HY-WU works in Multi-Agent System. Happy to collaborate with @yuzhang_shang @VITAGroupUT @huanwangx

[Xin Jin @Xander_K1ng]

@JulieKallini NICE WORK!!! BLT -> Fast LBT, awesome 🔥

[Xin Jin @Xander_K1ng]

@breadli428 @GoogleDeepMind awesome！chenhao🤩🤩🤩

[Xin Jin @Xander_K1ng]

Very enjoy sleeping with dragon.

[Sicheng Feng @si_feng32704]

[ICLR 2026] First time to attend top conference! Brazil!!!

[Liner @search_liner]

Liner is partnering with @spoticlr at #ICLR2026 — supporting Best Paper and Travel Awards for LLM research.

And to celebrate, we're giving away:
✈️ Round-trip flights + hotel to #ICML2026 in Seoul
🎁 $300 Liner Credits

Follow @search_liner + repost to enter by 4/27.

Liner is built for research workflows. Find papers, verify sources, and write with citations in one place.

See you in 🇧🇷 and 🇰🇷!

@iclr_conf @icmlconf

[Xin Jin @Xander_K1ng]

@TencentHunyuan excellent work🥳🥳 @VictorKaiWang1

[Tencent Hy @TencentHunyuan]

One static model does not fit all😭

We just dropped our latest work: Functional Neural Memory. Instead of static models, we generate custom "parameters" for every single input.

✅Prompt your model anytime
✅Instant personalization
✅Better instruction following
✅Flexible & dynamic memory (w/o memory bank✌️)

(🧵1/6)

[Xin Jin @Xander_K1ng]

@si_feng32704 Congratulations!

[Sicheng Feng @si_feng32704]

🚀 dVoting

We introduce dVoting, a fast voting strategy that boosts reasoning capability in dLLMs without any training.

- Training-free & Simple
- +7.66% on GSM8K, +7.20% on MATH500
- First efficient test-time scaling strategy for dLLMs

Page: https://t.co/QkysyekJn2

[Xin Jin @Xander_K1ng]

@Alright_lone NB-K, dear Junhan

[Xin Jin @Xander_K1ng]

@Alright_lone Thanks嘛？那很牛逼K了😡

[Xin Jin @Xander_K1ng]

Nice work！Junhan！The best undergraduate student in Westlake Uni🤩🤩

[Junhan Zhu @Alright_lone]

🚀 OBS-Diff

We introduce OBS-Diff: a novel, training-free, one-shot pruning framework for diffusion models, supporting diverse architectures and pruning granularities.

📄 Paper: https://t.co/RsvZHXXlwo
💻 Code: https://t.co/R6EIJXBMdh
🌐 Web: https://t.co/cQaX80tl3B

[Huan Wang @ICML26 @huanwangx]

#NeurIPS2025 Flying to SD to present our three papers on efficient and reliable AI. See you soon friends! ☺️

[Xin Jin @Xander_K1ng]

Nice work, jie ge 😎

[DU Wenjie @wjdu24]

💡Which Heads Matter for Reasoning? RL-Guided KV Cache Compression

We introduce RLKV, a novel reasoning-critical head identification method by reinforcement learning, to guide KV cache compression for reasoning LLMs.

Paper: https://t.co/Ng3rhkjBzU
Page: https://t.co/27FNGLdw8i

[Xin Jin @Xander_K1ng]

@Alright_lone @JinWeiyang18434 🤩

[Xin Jin @Xander_K1ng]

@OPatashnik @TelAvivUni Congrats!!!🥳🥳