Towards Verifiable Text Generation with Symbolic References

arXiv:2311.09188v2 Announce Type: replace-cross
Abstract: LLMs are vulnerable to hallucinations, and thus their outputs generally require laborious human verification for high-stakes applications. To this end, we propose symbolically grounded generation (SymGen) as a simple approach for enabling easier manual validation of an LLM's output. SymGen prompts an LLM to interleave its regular output text with explicit symbolic references to fields present in some conditioning data (e.g., a table in JSON format). The references can be used to display the provenance …

arxiv cs.ai cs.cl cs.lg text text generation type