Caiqi Zhang
@caiqizh
Followers
46
Following
9
Statuses
19
PhD student at University of Cambridge
Joined October 2023
RT @li_chengzu: Forget just thinking in words. š New Era of Multimodal ReasoningšØ š Imagine While Reasoning in Space with MVoT Multimodalā¦
0
163
0
RT @YinhongLiu2: šØ New Paper Alert! šØ When using LLMs for judgements, ever wondered about the consistency of those judgments? š¤ Check out oā¦
0
70
0
RT @ZhijiangG: Life update: š I'm excited to share that I will be joining @HKUSTGuangzhou as an Assistant Professor in Spring 2025! I'm loā¦
0
23
0
RT @ArtemShelmanov: I just met Caiqi Zhang ā one of the users of our Python framework for uncertainty quantification, LM-Polygraph. Itās inā¦
0
1
0
@ArtemShelmanov Nice to see you in the conference! LM-Polygraph is a really helpful framework!
0
0
2
š„ We also have some interesting preprint followup works. Check them out! - Atomic Calibration of LLMs in Long-Form Generations - LoGU: Long-form Generation with Uncertainty Expressions
0
0
2
I will present this work in EMNLP Nov 13 Wednesday! See you there!
Happy to share that TopViewRS is selected as oral presentation at @emnlpmeeting #EMNLP2024 Though Iām attending virtually, @caiqizh will present our work on Nov 13 11:45 to 12:00 at Ashe Auditorium. If you are interested in multimodality/spatial reasoning, feel free to reach out!
0
0
6
RT @li_chengzu: Happy to share that TopViewRS is selected as oral presentation at @emnlpmeeting #EMNLP2024 Though Iām attending virtually,ā¦
0
1
0
š„Check our EMNLP paper with @vlachos_nlp and @ZhijiangG š¤Do We Need Language-Specific Fact-Checking Models? The Case of Chinese ā¼ļø We find the domain and cultural biases in the Chinese fact-checking area that necessitate language-specific tools!
0
5
13
0
0
4
Key findings: - Models are more likely to conform when uncertain or facing difficult tasks *like humans*! š¤ - Instruction-tuned models are more resistant šŖ - Two simple interventions to reduce conformity: Devilās Advocate š and Question Distillationš§Ŗ!
0
0
4
RT @tiancheng_hu: Thrilled to share our new paper: "Can LLM be a Personalized Judge?" We investigate the reliability of LLMs in judging useā¦
0
22
0
RT @cambridgenlp: For this week's @cambridgenlp reading group, Xiaochen ( presented his recent paper "Conformity inā¦
0
4
0
RT @Zichen2501: Differentiable rendering made SIMPLEāļø Differentiating physically based renderers is hard: Dirac-delta discontinuities ariā¦
0
56
0
RT @Yulongchen1010: Evaluating LLMs usually requires sophisticated human designs and with the continuous improvement of LLMs, it is difficuā¦
0
24
0
RT @li_chengzu: To delve deeper into how models reason about space šø, I believe a lot more can be studied from cognitionš§ to embodimentš¤ļøbuiā¦
0
1
0