Paper Reading

Location: EE Old Committee Room, 3rd Floor, Block III, IIT Delhi.

29.08.2024: Arghodeep Nandi presented "UniT: A Unified Look at Certified Robust Training against Text Adversarial Perturbation"
29.08.2024: Sudipto Ghosh presented "UniversalNER: Targeted distillation from large language models for open-named entity recognition"
09.08.2024: Prottay Kumar Adhikary presented "LazyLLM: Dynamic Token Pruning for Efficient Long Context LLM Inference"
31.07.2024: Palash Nandi presented "Understanding Transformers via N-gram Statistics"
03.07.2024: Mohit Kataria and Sahil Mishra presented "GraphFlex: Scalable and Continuous Learning of Graph Structure."
06.03.2024: Eshaan Tanwar presented "Transformer and its variants."
17.01.2024: Anwoy Chatterjee and Aswini Kumar Padhi presented "Basics of RL (Policy Gradient, PPO, TRPO)."
10.01.2024: Aswini Kumar Padhi and Anwoy Chatterjee presented "Basics of RL (Q-Learning, Deep Q-Learning, PPO, TRPO)."
15.11.2023: Sahajpreet Singh presented "Interpreting Language Models with Contrastive Explanations."
08.11.2023: Shubhanshu Khatana presented "AMPNet: Attention as Message Passing for Graph Neural Networks."
01.11.2023: Pooja Singh presented "Graph Generative Model for Benchmarking Graph Neural Networks."
11.10.2023: Shaonli Pal presented "Cooperative Graph Neural Networks."
04.10.2023: Anik Chatterjee presented "RLPROMPT: Optimizing Discrete Text Prompts with Reinforcement Learning."
27.09.2023: Aswini Kumar Padhi presented "Faith and Fate: Limits of Transformers on Compositionality."
13.09.2023: Ayan Sengupta presented "What Can Transformers Learn In-Context? A Case Study of Simple Function Classes."
06.09.2023: Palash Nandi presented "Verify-and-Edit: A Knowledge-Enhanced Chain-of-Thought Framework."
30.08.2023: Prasoon Bajpai presented "Graph of Thoughts: Solving Elaborate Problems with Large Language Models."
23.08.2023: Anwoy Chatterjee presented "Z-ICL: Zero-Shot In-Context Learning with Pseudo-Demonstrations."
16.08.2023: Amey Hengle presented "Diffusion-LM Improves Controllable Text Generation."
09.08.2023: Eshaan Tanwar presented "TEMPERA: Test-Time Prompting via Reinforcement Learning."