ici-mt / reading-list

A paper reading list maintained by ICI-MT, contains all papers and PPTs (if available) shared by our groups.

Reading-List

A paper reading list maintained by ICI-MT, contains all papers and PPTs (if available) shared by our groups.

2023/9/04

• A Survey on Evaluation of Large Language Models
• Do Androids Laugh at Electric Sheep? Humor "Understanding" Benchmarks from The New Yorker Caption Contest, ACL 2023 Best Papers

2023/10/13

• Entity Tracking in Language Models, ACL 2023, PPT

2023/11/23

• Evaluating Large Language Models at Evaluating Instruction Following, ICLR 2024, PPT
• Presented in [2023.11.22] 文献阅读-李泽林.pdf:
  • LLM Lies: Hallucinations are not Bugs, but Features as Adversarial Examples, ICLR 2024 (5, 3, 3, 3);
  • Who’s Harry Potter? Approximate Unlearning for LLMs, ICLR 2024 (8, 3, 5)
  • AutoDAN: Automatic and Interpretable Adversarial Attacks on Large Language Models, ICLR 2024 (5, 5, 10, 5).

2023/12/07

• SALMONN: TOWARDS GENERIC HEARING ABILITIES FOR LARGE LANGUAGE MODELS, ICLR 2024 (8, 6, 6), PPT

2023/12/14

• Large Language Models are Efficient Learners of Noise-Robust Speech Recognition, ICLR 2024 (8, 10, 6, 8), PPT

2023/12/21

• FUNCTION VECTORS IN LARGE LANGUAGE MODELS, arXiv Oct 23 2023, PPT.
• Label Words are Anchors: An Information Flow Perspective for Understanding In-Context Learning, EMNLP 2023 Best Paper, PPT.
• Graph-Guided Reasoning for Multi-Hop Question Answering, PPT. ,

2023/12/28

• Adapting Language Models to Compress Contexts, EMNLP 2023, PPT
• Correction with Backtracking Reduces Hallucination in Summarization, [cs.CL],PPT

2024/03/19

• In-context Vectors: Making In Context Learning More Effective and Controllable Through Latent Space Steering arXiv Nov. 11 2023, Presentation