Tingchen Fu
@TingchenFu
Followers
214
Following
1K
Media
13
Statuses
182
Incoming PhD student @UniofOxford and @MetaAI, prev Renmin University of China (RUC)
Beijing, China
Joined September 2022
1/🧵 New research alerted🚨: "Scaling Reasoning, Losing Control: Evaluating Instruction Following in Large Reasoning Models" (Paper:. We found a surprising paradox between instruction-following ability and reasoning ability. Here’s why ↓
2
3
31
RT @OwainEvans_UK: New paper & surprising result. LLMs transmit traits to other models via hidden signals in data. Datasets consisting only….
0
1K
0
RT @micahgoldblum: 🚨 Did you know that small-batch vanilla SGD without momentum (i.e. the first optimizer you learn about in intro ML) is v….
0
106
0
RT @rohanpaul_ai: the Grok 4 benchmark chart (leaked version) is just beautiful. Did @xAI really hit 45% on HLE (Humanities Last Exam) 🤯. B….
0
7
0
RT @valentina__py: 💡Beyond math/code, instruction following with verifiable constraints is suitable to be learned with RLVR. But the set of….
0
84
0
RT @yupenghou97: All reviews: positive; Meta: accept; Rejected by #RecSys2025. I get that decisions are complex, eg, "maintaining a competi….
0
6
0
RT @FazlBarez: Excited to share our paper: "Chain-of-Thought Is Not Explainability"! . We unpack a critical misconception in AI: models exp….
0
133
0
RT @Benjamin_eecs: We've always been excited about self-play unlocking continuously improving agents. Our insight: RL selects generalizable….
0
52
0
RT @fengyao1909: 😵💫 Struggling with 𝐟𝐢𝐧𝐞-𝐭𝐮𝐧𝐢𝐧𝐠 𝐌𝐨𝐄?. Meet 𝐃𝐞𝐧𝐬𝐞𝐌𝐢𝐱𝐞𝐫 — an MoE post-training method that offers more 𝐩𝐫𝐞𝐜𝐢𝐬𝐞 𝐫𝐨𝐮𝐭𝐞𝐫 𝐠𝐫𝐚𝐝𝐢𝐞….
0
55
0
RT @rohanpaul_ai: 🚨 CHINA’S BIGGEST PUBLIC AI DROP SINCE DEEPSEEK. @Baidu_Inc open source Ernie, 10 multimodal MoE variants. 🔥 Surpasses De….
0
97
0
RT @_andreilupu: Theory of Mind (ToM) is crucial for next gen LLM Agents, yet current benchmarks suffer from multiple shortcomings. Enter….
0
30
0
RT @jpineau1: I'm excited to be joining the board of the Laude Institute! We need more support and incentives for university researchers….
0
21
0
RT @JonnyCoook: Can an LLM be programmed?. In our new preprint, we show that LLMs can learn to evaluate programs for a range of inputs by b….
0
33
0
RT @iScienceLuvr: EvoLM: In Search of Lost Language Model Training Dynamics. "We present EvoLM, a model suite that enables systematic and t….
0
25
0
RT @HuggingPapers: Discrete Diffusion in Large Language and Multimodal Models: A Survey just released on Hugging Face. Get an overview of r….
0
94
0
RT @AnChancy46881: # 🚨 4B open-recipe model beats Claude-4-Opus .🔓 100% open data, recipe, model weights and code. Introducing Polaris✨--a….
0
81
0
RT @aclmeeting: 🚨 Just a reminder! 🚨 The Main Conference Papers for ACL 2025 are now available! 🤩 Dive into the cutting-edge research being….
2025.aclweb.org
ACL 2025 Accepted Main Conference Papers
0
23
0