Open Issues Need Help
View All on GitHub Add a benchmark whose tasks branch into several lines of inquiry, as an evaluation-only suite about 1 month ago
good first issue
Asking for What Was Never Requested: horizontal and vertical proactivity in LLM agents. Need-graph metrics (no LLM judge) and Q&D, which trains a questioner from the consequences of its questions.
Python
#agent-evaluation#clarifying-questions#customer-service#dpo#huggingface#information-seeking#large-language-models#llm-agents#lora#machine-learning#multi-hop-qa#nlp#peft#preference-learning#proactive-agents#question-asking#qwen3#rag#retrieval-augmented-generation#tau-bench
Retrieval pools where a prerequisite must be found before its dependents about 1 month ago
enhancement help wanted
Asking for What Was Never Requested: horizontal and vertical proactivity in LLM agents. Need-graph metrics (no LLM judge) and Q&D, which trains a questioner from the consequences of its questions.
Python
#agent-evaluation#clarifying-questions#customer-service#dpo#huggingface#information-seeking#large-language-models#llm-agents#lora#machine-learning#multi-hop-qa#nlp#peft#preference-learning#proactive-agents#question-asking#qwen3#rag#retrieval-augmented-generation#tau-bench
Stop supervision that carries the evidence gain into final answers about 1 month ago
enhancement help wanted
Asking for What Was Never Requested: horizontal and vertical proactivity in LLM agents. Need-graph metrics (no LLM judge) and Q&D, which trains a questioner from the consequences of its questions.
Python
#agent-evaluation#clarifying-questions#customer-service#dpo#huggingface#information-seeking#large-language-models#llm-agents#lora#machine-learning#multi-hop-qa#nlp#peft#preference-learning#proactive-agents#question-asking#qwen3#rag#retrieval-augmented-generation#tau-bench