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
20 Pith papers cite this work. Polarity classification is still indexing.
20
Pith papers citing it
representative citing papers
QAOD projects away question-aligned directions from answer representations to isolate domain-agnostic factuality signals, enabling efficient hallucination detection with top in-domain AUROC and up to 21% better OOD transfer.
LOVER creates an unsupervised logic-regularized verifier that reaches 95% of supervised verifier performance on reasoning tasks across 10 datasets.
RouterBench supplies a standardized benchmark, 405k+ inference dataset, theoretical framework, and comparative analysis for multi-LLM routing systems.
PGT generates synthetic tasks via geometric overlays on images to supply dense visual supervision, improving spatial and relational understanding in MLLMs by up to 20% on targeted benchmarks.
A dual hierarchical RL framework with two agents coordinates high-level dialogue strategy and low-level question generation to emulate judicial questioning and extract key information from Supreme Court arguments, outperforming baselines.
DKPS-based methods leverage cached model responses to achieve equivalent benchmark prediction accuracy with substantially fewer queries than standard evaluation.
A blind replay script matches frontier model performance on static CUA benchmarks due to non-principled environments and evaluation methods, prompting PRISM design principles and the DigiWorld benchmark with improved statistical aggregation.
BioTool dataset enables fine-tuning a 4B-parameter LLM to outperform GPT-5.1 in biomedical tool calling while improving downstream answer quality per human experts.
ZAYA1-8B is a reasoning MoE model with 700M active parameters that matches larger models on math and coding benchmarks and reaches 91.9% on AIME'25 via Markovian RSA test-time compute.
ClawEnvKit automates generation of diverse verified environments for claw-like agents from natural language, producing the Auto-ClawEval benchmark of 1,040 environments that matches human-curated quality at 13,800x lower cost.
A two-phase non-parametric retrieval workflow that separates high-recall candidate retrieval from high-precision utility ranking with LLM-as-a-Judge scoring for evidence extraction in multilingual financial documents.
GS-Quant generates coarse-to-fine discrete codes for KG entities via semantic hierarchy injection and causal sequence reconstruction, enabling LLMs to perform knowledge graph completion by treating the codes as vocabulary tokens.
RAG is more effective and cost-efficient than fine-tuning for industrial QA adaptation on automotive datasets.
Gemma introduces open 2B and 7B LLMs derived from Gemini technology that beat comparable open models on 11 of 18 text tasks and come with safety assessments.
A case-driven multi-agent system automates the full pipeline of bad-case detection, annotation, and resolution for e-commerce search relevance using Annotator, Optimizer, and User agents plus supporting components.
Gemma 2 models achieve leading performance at their sizes by combining established Transformer modifications with knowledge distillation for the 2B and 9B variants.
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.
-
When Answers Stray from Questions: Hallucination Detection via Question-Answer Orthogonal Decomposition
QAOD projects away question-aligned directions from answer representations to isolate domain-agnostic factuality signals, enabling efficient hallucination detection with top in-domain AUROC and up to 21% better OOD transfer.
-
Logic-Regularized Verifier Elicits Reasoning from LLMs
LOVER creates an unsupervised logic-regularized verifier that reaches 95% of supervised verifier performance on reasoning tasks across 10 datasets.
-
RouterBench: A Benchmark for Multi-LLM Routing System
RouterBench supplies a standardized benchmark, 405k+ inference dataset, theoretical framework, and comparative analysis for multi-LLM routing systems.
-
PGT: Procedurally Generated Tasks for improving visual grounding in MLLMs
PGT generates synthetic tasks via geometric overlays on images to supply dense visual supervision, improving spatial and relational understanding in MLLMs by up to 20% on targeted benchmarks.
-
Dual Hierarchical Dialogue Policy Learning for Legal Inquisitive Conversational Agents
A dual hierarchical RL framework with two agents coordinates high-level dialogue strategy and low-level question generation to emulate judicial questioning and extract key information from Supreme Court arguments, outperforming baselines.
-
Query-efficient model evaluation using cached responses
DKPS-based methods leverage cached model responses to achieve equivalent benchmark prediction accuracy with substantially fewer queries than standard evaluation.
-
Computer Use at the Edge of the Statistical Precipice
A blind replay script matches frontier model performance on static CUA benchmarks due to non-principled environments and evaluation methods, prompting PRISM design principles and the DigiWorld benchmark with improved statistical aggregation.
-
BioTool: A Comprehensive Tool-Calling Dataset for Enhancing Biomedical Capabilities of Large Language Models
BioTool dataset enables fine-tuning a 4B-parameter LLM to outperform GPT-5.1 in biomedical tool calling while improving downstream answer quality per human experts.
-
ZAYA1-8B Technical Report
ZAYA1-8B is a reasoning MoE model with 700M active parameters that matches larger models on math and coding benchmarks and reaches 91.9% on AIME'25 via Markovian RSA test-time compute.
-
ClawEnvKit: Automatic Environment Generation for Claw-Like Agents
ClawEnvKit automates generation of diverse verified environments for claw-like agents from natural language, producing the Auto-ClawEval benchmark of 1,040 environments that matches human-curated quality at 13,800x lower cost.
-
Beyond Semantic Similarity: A Two-Phase Non-Parametric Retrieval Workflow for Corporate Credit Underwriting
A two-phase non-parametric retrieval workflow that separates high-recall candidate retrieval from high-precision utility ranking with LLM-as-a-Judge scoring for evidence extraction in multilingual financial documents.
-
GS-Quant: Granular Semantic and Generative Structural Quantization for Knowledge Graph Completion
GS-Quant generates coarse-to-fine discrete codes for KG entities via semantic hierarchy injection and causal sequence reconstruction, enabling LLMs to perform knowledge graph completion by treating the codes as vocabulary tokens.
-
Assessment of RAG and Fine-Tuning for Industrial Question-Answering-Applications
RAG is more effective and cost-efficient than fine-tuning for industrial QA adaptation on automotive datasets.
-
Gemma: Open Models Based on Gemini Research and Technology
Gemma introduces open 2B and 7B LLMs derived from Gemini technology that beat comparable open models on 11 of 18 text tasks and come with safety assessments.
-
A Case-Driven Multi-Agent Framework for E-Commerce Search Relevance
A case-driven multi-agent system automates the full pipeline of bad-case detection, annotation, and resolution for e-commerce search relevance using Annotator, Optimizer, and User agents plus supporting components.
-
Gemma 2: Improving Open Language Models at a Practical Size
Gemma 2 models achieve leading performance at their sizes by combining established Transformer modifications with knowledge distillation for the 2B and 9B variants.
- Weasel: Out-of-Domain Generalization for Web Agents via Importance-Diversity Data Selection
- STABLEVAL: Disagreement-Aware and Stable Evaluation of AI Systems
- Lessons from the Trenches on Reproducible Evaluation of Language Models