Issues

• potential issue in get_sentence_embedding()

  Princeton-SysML/Jailbreak_LLM#13 · Filter by author terarachang opened on May 31, 2024

• For llama2, use_default config seems to have top-k values 50 because of the default value of hugging face config, but did llama2 use top-k?

  Princeton-SysML/Jailbreak_LLM#12 · Filter by author along-l opened on May 4, 2024
  1

• Did system prompts are used for GCG and Generation Exploitation methods respectively in experimental results reported in the paper?

  Princeton-SysML/Jailbreak_LLM#11 · Filter by author along-l opened on May 2, 2024
  1

• What is the difference between top_k=1 and greedy decoding, and why should we experiment separately?

  Princeton-SysML/Jailbreak_LLM#10 · Filter by author zggg1p opened on Feb 26, 2024
  1

• The output of the mpt-30b-chat contains a large number of irrelevant characters

  Princeton-SysML/Jailbreak_LLM#9 · Filter by author zggg1p opened on Feb 17, 2024

• Missing Chat Template

  Princeton-SysML/Jailbreak_LLM#8 · Filter by author justinphan3110cais opened on Jan 31, 2024
  2

• Issues while Running FlanT5-small model with JailBreak

  Princeton-SysML/Jailbreak_LLM#7 · Filter by author aneekroy opened on Dec 22, 2023

• Another Question about Default Transformer Decoding Config

  clarification
  Princeton-SysML/Jailbreak_LLM#5 · Filter by author chujiezheng opened on Nov 24, 2023
  1

• About Input Formats for Different Models

  ablation
  Princeton-SysML/Jailbreak_LLM#4 · Filter by author chujiezheng opened on Nov 23, 2023
  3

• (Maybe) Bugs in the code.

  bug
  Princeton-SysML/Jailbreak_LLM#3 · Filter by author Junjie-Chu opened on Nov 21, 2023
  1