|
Workshop
KCC XAI
28 June 2024 [pdf] room 1805 ๐ |
Representation Interpretation of Refusal Mechanism In Large Language ModelsWe adapt mechanistic interpretability on the refusal mechanism by proposing Prompt and Location-based Refusal Analysis (PLoRA), which compares refusal feature vectors generated from the specific prompts and at the locations of tokens. |
|
|
Conference
IJCAI
09 Aug 2024 [project] [arxiv] |
Memorizing Documents with Guidance in Large Language ModelsWe propose document-wise memories, which makes document-wise locations for neural memories in training. The proposed architecture maps document representation to memory entries and filters memory selections in the forward process of LLMs. |
|
|
Conference
ICPRAI
12 June 2024
[project] |
Identifying the Source of Generation for Large Language ModelsThis work introduces token-level source identification in the decoding step, which maps the token representation to the reference document. We propose a bi-gram source identifier which has two successive token representations as input. |
|