←
LCS2 Lab762 posts
LCS2 Lab avatar

LCS2 Lab

@lcs2lab

Lab. for Computational Social Systems, a group led by @Tanmoy_Chak working on #SocialComputing #GraphMining & #NLProc

◷ Joined July 2018▧ 210 Photos and videos
LCS2 Lab avatar
LCS2 Lab@lcs2lab·Nov 17

🏆 Prottay Kumar Adhikary — Winner, Demonstrations Title: coTherapist: Intelligent Support System for Mental Healthcare Experts Thematic area: Health, Biotechnology and Medical Devices

Post attachment
LCS2 Lab avatar
LCS2 Lab@lcs2lab·Nov 17

🏆 Indraayudh Talukdar — Winner, Posters Title: Meta-Evaluation of LLM-as-a-Judge Thematic area: AI, Machine Learning & Data Science

Post attachment
Tanmoy Chakraborty avatar
⟳ LCS2 Lab retweeted
Tanmoy Chakraborty@Tanmoy_Chak·Nov 11

Delighted to announce TRUSTMAPS -- an @ICMRDELHI -funded National Centre led by @iitdelhi with AIIMS Delhi, IIT Kharagpur, Ashoka University & Tata Medical Centre. As a PI, I will work with 18 Co-PIs across 5 institutes. Our mission: Build India's first national certification ecosystem for Medical AI. More details here: lnkd.in/p/g9gi7yNY 📢 Follow the official TRUSTMAPS LinkedIn page for research updates, collaborations, and upcoming initiatives: linkedin.com/company/trustma… @AshokaUniv @aiims_newdelhi @IITKgp

Post attachment
LCS2 Lab avatar
LCS2 Lab@lcs2lab·Nov 4

🚀 New Paper Alert 🚀 Excited to share that our paper, “Counter with peace! Persona-Guided Discrete Stochastic Processes for Conditioned Counterspeech Generation,” has been accepted to the Findings Track of #WSDM2027! 🎉 👥 @aswiniandpunter, Apoorv Jain, @Tanmoy_Chak

Post attachment
LCS2 Lab avatar
LCS2 Lab@lcs2lab·Oct 29

✅ HallusionBench: AUROC rises from 0.634 to 0.769. ✅ VizWiz: AUROC rises from 0.707 to 0.817. ✅ Pooled XGBoost: AUROC improves from 0.78 with scalar confidence to 0.95 with the three diagnostic signals and 0.97 when confidence is added.

LCS2 Lab avatar
LCS2 Lab@lcs2lab·Oct 29

HalluPrism probes model responses under controlled perturbations to construct a three-dimensional Perturbation-Response Signature to study how answers behave when visual evidence is degraded, removed, or challenged not just how confident the model sounds.

LCS2 Lab avatar
LCS2 Lab@lcs2lab·Oct 28

✅ Experts preferred MAGneT-generated sessions in 77.2% of cases on average across nine dimensions. ✅ Generated sessions achieved 3.2% higher general counseling skills and 4.3% higher CBT-specific skills on CTRS than the strongest baseline.

LCS2 Lab avatar
LCS2 Lab@lcs2lab·Oct 28

⚙️A CBT agent develops a session-level plan, while a technique-selection agent dynamically selects appropriate techniques for each turn. A response generator then integrates the agents’ outputs into a coherent counselor response.

LCS2 Lab avatar
LCS2 Lab@lcs2lab·Oct 28

🧠 Counseling requires diverse psychological techniques, appropriate strategy selection, and consistency across multiple turns. MAGneT decomposes counselor response generation into specialized agents: Reflection, Questioning, Solution provision, Normalization, Psycho-education.