ORPO performs preference alignment during supervised fine-tuning via a monolithic odds ratio penalty, allowing 7B models to outperform larger state-of-the-art models on alignment benchmarks.
Title resolution pending
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
Agent Island is a new multiagent game environment that functions as a dynamic benchmark resistant to saturation and contamination, with Bayesian ranking showing OpenAI GPT-5.5 as the strongest performer among 49 models across 999 games.
A total-variation robust Plackett-Luce loss whose adversarial ranking is the ascending-score order, with convex offline and weakly-convex online guarantees.
ORM-based test-time verification improves Text-to-SQL accuracy over heuristic selection by up to 4.33% on BIRD and 2.10% on Spider using automated labeling.
LLM embeddings condition a generative transformer to enable faster convergence, better performance, and generalization to unseen LHC processes using a single model.
Risk-sensitive preference games using convex risk measures produce policies that are robust across data strata and match or exceed standard Nash learning performance without added cost.
citing papers explorer
-
ORPO: Monolithic Preference Optimization without Reference Model
ORPO performs preference alignment during supervised fine-tuning via a monolithic odds ratio penalty, allowing 7B models to outperform larger state-of-the-art models on alignment benchmarks.
-
Agent Island: A Saturation- and Contamination-Resistant Benchmark from Multiagent Games
Agent Island is a new multiagent game environment that functions as a dynamic benchmark resistant to saturation and contamination, with Bayesian ranking showing OpenAI GPT-5.5 as the strongest performer among 49 models across 999 games.
-
Distributionally Robust Listwise Preference Optimization
A total-variation robust Plackett-Luce loss whose adversarial ranking is the ascending-score order, with convex offline and weakly-convex online guarantees.
-
Test-Time Verification for Text-to-SQL via Outcome Reward Models
ORM-based test-time verification improves Text-to-SQL accuracy over heuristic selection by up to 4.33% on BIRD and 2.10% on Spider using automated labeling.
-
One Generator, Any Process: LLM-Conditioning for the LHC
LLM embeddings condition a generative transformer to enable faster convergence, better performance, and generalization to unseen LHC processes using a single model.
-
Structure from Strategic Interaction & Uncertainty: Risk Sensitive Games for Robust Preference Learning
Risk-sensitive preference games using convex risk measures produce policies that are robust across data strata and match or exceed standard Nash learning performance without added cost.
- Structuring license permissiveness from pairwise comparisons