List your company

Hallucination Detection on NQ (test)

95.2AUC ROCCDES
40.91255.00669.183.194May 30, 2025Jul 25, 2025Sep 20, 2025Nov 16, 2025Jan 11, 2026Mar 9, 2026May 5, 2026

Evaluation results

MethodLinks
CDESLLM Backbone=Llama-2-13B2025.0595.2——————
SDESLLM Backbone=Llama-2-7B2025.0594.1——————
CDESLLM Backbone=Llama-2-7B2025.0593.7——————
MINDLLM Backbone=Llama-2-13B2025.0593.6——————
CDESLLM Backbone=Vicuna-13B2025.0593.6——————
SDESLLM Backbone=Alpaca-13B2025.0593.5——————
SAPLMALLM Backbone=Vicuna-13B2025.0593.2——————
SDESLLM Backbone=Llama-2-13B2025.0593.2——————
Probe@ExactLLM Backbone=Alpaca-13B2025.0592.8——————
MINDLLM Backbone=Alpaca-13B2025.0592.7——————
ODESLLM Backbone=Vicuna-13B2025.0592.4——————
Probe@ExactLLM Backbone=Vicuna-13B2025.0592.3——————
ODESLLM Backbone=Llama-2-13B2025.0592.1——————
CDESLLM Backbone=Alpaca-13B2025.0592.1——————
Probe@ExactLLM Backbone=Llama-2-7B2025.0592——————
Probe@ExactLLM Backbone=Llama-2-13B2025.0591.9——————
CDESLLM Backbone=Gemma-2-9B2025.0591.8——————
ODESLLM Backbone=Llama-2-7B2025.0591.3——————
SDESLLM Backbone=Vicuna-13B2025.0591.1——————
SDESLLM Backbone=Gemma-2-9B2025.0590.9——————
MINDLLM Backbone=Vicuna-13B2025.0590.6——————
SAPLMALLM Backbone=Llama-2-13B2025.0590.5——————
MINDLLM Backbone=Llama-2-7B2025.0590.5——————
ODESLLM Backbone=Alpaca-13B2025.0590.5——————
CDESLLM Backbone=Mistral-7B-v0.32025.0590.5——————
SAPLMALLM Backbone=Mistral-7B-v0.32025.0590.3——————
Probe@ExactLLM Backbone=Gemma-2-9B2025.0590.3——————
ODESLLM Backbone=Gemma-2-9B2025.0590——————
ODESLLM Backbone=Mistral-7B-v0.32025.0589.7——————
SDESLLM Backbone=Mistral-7B-v0.32025.0589.7——————
MINDLLM Backbone=Gemma-2-9B2025.0589.5——————
SAPLMALLM Backbone=Llama-2-7B2025.0589.1——————
SAPLMALLM Backbone=Gemma-2-9B2025.0588.9——————
Probe@ExactLLM Backbone=Mistral-7B-v0.32025.0588.6——————
SAPLMALLM Backbone=Alpaca-13B2025.0587.6——————
MINDLLM Backbone=Mistral-7B-v0.32025.0587.2——————
TDGNetgraph_type=dynamic, decoding_gen_length=32, decoding_steps=16, granularity=token-level2026.0287——————
CHARMgraph_type=static, decoding_gen_length=32, decoding_steps=16, granularity=token-level2026.0284——————
ACSELLM Backbone=Mistral-7B-Instruct2026.0584——36298187
CAPLLM Backbone=Mistral-7B-Instruct2026.0580——50427877
EUBHDLLM Backbone=Llama-2-13B2025.0579.4——————
DDCRP-CPLLM Backbone=Mistral-7B-Instruct2026.0578——54467575
INSIDELLM Backbone=Llama-2-13B2025.0577.1——————
EUBHDLLM Backbone=Alpaca-13B2025.0576.8——————
INSIDELLM Backbone=Llama-2-7B2025.0576.8——————
INSIDELLM Backbone=Mistral-7B-v0.32025.0576.4——————
INSIDELLM Backbone=Gemma-2-9B2025.0575.9——————
INSIDELLM Backbone=Vicuna-13B2025.0575.8——————
SULLM Backbone=Mistral-7B-Instruct2026.0575——58507273
INSIDELLM Backbone=Alpaca-13B2025.0574.3——————
EigVLLM Backbone=Mistral-7B-Instruct2026.0574——66597069
EUBHDLLM Backbone=Llama-2-7B2025.0573.9——————
EUBHDLLM Backbone=Vicuna-13B2025.0573.2——————
EUBHDLLM Backbone=Mistral-7B-v0.32025.0571.7——————
EUBHDLLM Backbone=Gemma-2-9B2025.0570.3——————
Predictive Entropydecoding_gen_length=32, decoding_steps=16, granularity=token-level2026.0269——————
UnigramLLM Backbone=Vicuna-13B2025.0567.8——————
P(True)LLM Backbone=Mistral-7B-Instruct2026.0567——74666464
NLILLM Backbone=Llama-2-13B2025.0566.9——————
NLILLM Backbone=Vicuna-13B2025.0565.4——————
UnigramLLM Backbone=Llama-2-13B2025.0565.2——————
UnigramLLM Backbone=Gemma-2-9B2025.0565.1——————
UnigramLLM Backbone=Mistral-7B-v0.32025.0564.5——————
NLILLM Backbone=Llama-2-7B2025.0564.2——————
NLILLM Backbone=Gemma-2-9B2025.0564——————
NLILLM Backbone=Alpaca-13B2025.0563.8——————
UnigramLLM Backbone=Llama-2-7B2025.0563.1——————
UnigramLLM Backbone=Alpaca-13B2025.0562.9——————
NLILLM Backbone=Mistral-7B-v0.32025.0562.6——————
TELLM Backbone=Mistral-7B-Instruct2026.0562——82755760
Token Probabilitydecoding_gen_length=32, decoding_steps=16, granularity=token-level2026.0260——————
P(True)LLM Backbone=Llama-2-13B2025.0556.8——————
AvgProbLLM Backbone=Vicuna-13B2025.0556.4——————
AvgProbLLM Backbone=Llama-2-13B2025.0555.9——————
AvgProbLLM Backbone=Gemma-2-9B2025.0555.3——————
AvgEntLLM Backbone=Vicuna-13B2025.0555.2——————
P(True)LLM Backbone=Llama-2-7B2025.0554.7——————
AvgProbLLM Backbone=Mistral-7B-v0.32025.0554.7——————
AvgEntLLM Backbone=Llama-2-13B2025.0554.6——————
AvgEntLLM Backbone=Gemma-2-9B2025.0554.6——————
AvgProbLLM Backbone=Llama-2-7B2025.0554.3——————
AvgEntLLM Backbone=Alpaca-13B2025.0554.2——————
AvgEntLLM Backbone=Llama-2-7B2025.0553.9——————
AvgEntLLM Backbone=Mistral-7B-v0.32025.0553.8——————
P(True)LLM Backbone=Vicuna-13B2025.0553.4——————
AvgProbLLM Backbone=Alpaca-13B2025.0553.1——————
P(True)LLM Backbone=Mistral-7B-v0.32025.0552.1——————
Source Attributiondecoding_gen_length=32, decoding_steps=16, granularity=token-level2026.0252——————
P(True)LLM Backbone=Gemma-2-9B2025.0551.3——————
P(True)LLM Backbone=Alpaca-13B2025.0551——————
Graph Degree (Centrality)decoding_gen_length=32, decoding_steps=16, granularity=token-level2026.0243——————
CRGn_cal=595, Target=95%2025.12—95.80————