Introduces RuVerBench benchmark showing frontier LLMs achieve strong but noisy performance on rubric verification for agentic scenarios, with analysis of prompt, batching, and voting strategies.
hub
Alpagasus: Training a better alpaca with fewer data.arXiv preprint arXiv:2307.08701
21 Pith papers cite this work. Polarity classification is still indexing.
21
Pith papers citing it
hub tools
citation-role summary
background 4
citation-polarity summary
verdicts
UNVERDICTED 21roles
background 4polarities
background 4representative citing papers
CodeBlock partitions code responses into syntactically coherent blocks, scores them with generalized cross-entropy and data-flow signals, and applies sparse supervision to achieve higher pass@1 than full SFT using 1.9% of tokens on six benchmarks.
PII can be reconstructed from SFT models via prefix attacks, with the new COVA algorithm improving success rates and leakage varying by attacker knowledge and PII type.
A pair selection strategy based on negative similarity dynamics strengthens contrastive supervision in gloss-free sign language translation by reducing noisy negatives.
Magpie synthesizes 300K high-quality alignment instructions from Llama-3-Instruct via auto-regressive prompting on partial templates, enabling fine-tuned models to match official instruct performance on AlpacaEval, ArenaHard, and WildBench.
HallusionBench shows GPT-4V reaches only 31.42% accuracy on paired questions testing language hallucination and visual illusion in LVLMs, with other models below 16%.
SHIFT selects compact RLVR training subsets using the magnitude of hidden-state change from a single inference rollout plus quality-weighted farthest-first coverage, outperforming training-free baselines on math reasoning and medical QA under low budgets.
Characterizes spurious correlation mechanisms in preference optimization via mean spurious bias and causal-spurious correlation leakage, demonstrates irreducible vulnerability to distribution shift, and introduces tie training as selective mitigation with validation on log-linear models and empirica
Target-aligned data selection via normalized endpoint loss drop on a validation-induced reference path achieves competitive performance with reduced computational overhead.
Introduces VURB benchmark and VUP-35K dataset to train discriminative and generative video reward models that achieve SOTA performance on VURB and VideoRewardBench.
HEAL mitigates entropy collapse in few-shot RLVR by selectively adding general-domain data and aligning trajectory-level entropy dynamics, matching full-shot performance with 32 target samples.
LLM agents iteratively generate and optimize data processing strategies for fine-tuning, delivering over 80% win rates versus unprocessed data and 65% versus LLM-based AutoML baselines while cutting search time by up to 10x.
Training-Trajectory-Aware Token Selection (T3S) reconstructs the token-level training objective to overcome a performance bottleneck in continual distillation of reasoning capabilities from large to small language models.
SymNoise applies symmetric noise to embeddings during instruction fine-tuning and reports 6.7% higher AlpacaEval scores than NEFTune on LLaMA-2-7B.
Cost-Aware SGD samples by gradient-norm-to-cost ratio and is instantiated as Cost-Aware GRPO for length-dependent policy gradients, reducing tokens used in LLM RL while matching baseline accuracy.
OptProver transfers formal theorem proving from Olympiad math to optimization via continual training, achieving SOTA Pass@1 and Pass@32 on a new Lean 4 benchmark while retaining general performance.
ADAPT is an online reweighting framework for LLM training that outperforms offline data selection and mixing methods in cross-benchmark generalization under equal compute.
MM-LIMA uses proposed quality metrics and a trainable selector to pick 200 high-quality multimodal instruction examples and outperforms MiniGPT-4 on evaluations.
Proxy RL produces a staged proxy-internalization capability that emerges before and predicts reward hacking in coding environments.
XekRung achieves state-of-the-art performance on cybersecurity benchmarks among same-scale models via tailored data synthesis and multi-stage training while retaining strong general capabilities.
A survey of positional encoding methods in transformer-based time series models that evaluates fixed, learnable, relative, and hybrid approaches on classification tasks and links effectiveness to data characteristics.
citing papers explorer
-
Can LLM-as-a-Judge Reliably Verify Rubrics in Agentic Scenarios?
Introduces RuVerBench benchmark showing frontier LLMs achieve strong but noisy performance on rubric verification for agentic scenarios, with analysis of prompt, batching, and voting strategies.
-
CODEBLOCK: Learning to Supervise Code at the Right Granularity
CodeBlock partitions code responses into syntactically coherent blocks, scores them with generalized cross-entropy and data-flow signals, and applies sparse supervision to achieve higher pass@1 than full SFT using 1.9% of tokens on six benchmarks.
-
Reconstruction of Personally Identifiable Information from Supervised Finetuned Models
PII can be reconstructed from SFT models via prefix attacks, with the new COVA algorithm improving success rates and leakage varying by attacker knowledge and PII type.
-
Selective Contrastive Learning For Gloss Free Sign Language Translation
A pair selection strategy based on negative similarity dynamics strengthens contrastive supervision in gloss-free sign language translation by reducing noisy negatives.
-
Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing
Magpie synthesizes 300K high-quality alignment instructions from Llama-3-Instruct via auto-regressive prompting on partial templates, enabling fine-tuned models to match official instruct performance on AlpacaEval, ArenaHard, and WildBench.
-
HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Models
HallusionBench shows GPT-4V reaches only 31.42% accuracy on paired questions testing language hallucination and visual illusion in LVLMs, with other models below 16%.
-
Single-Rollout Hidden-State Dynamics for Training-Free RLVR Data Selection
SHIFT selects compact RLVR training subsets using the magnitude of hidden-state change from a single inference rollout plus quality-weighted farthest-first coverage, outperforming training-free baselines on math reasoning and medical QA under low budgets.
-
Spurious Correlation Learning in Preference Optimization: Mechanisms, Consequences, and Mitigation via Tie Training
Characterizes spurious correlation mechanisms in preference optimization via mean spurious bias and causal-spurious correlation leakage, demonstrates irreducible vulnerability to distribution shift, and introduces tie training as selective mitigation with validation on log-linear models and empirica
-
Let the Target Select for Itself: Data Selection via Target-Aligned Paths
Target-aligned data selection via normalized endpoint loss drop on a validation-induced reference path achieves competitive performance with reduced computational overhead.
-
Video Understanding Reward Modeling: A Robust Benchmark and Performant Reward Models
Introduces VURB benchmark and VUP-35K dataset to train discriminative and generative video reward models that achieve SOTA performance on VURB and VideoRewardBench.
-
HEALing Entropy Collapse: Enhancing Exploration in Few-Shot RLVR via Hybrid-Domain Entropy Dynamics Alignment
HEAL mitigates entropy collapse in few-shot RLVR by selectively adding general-domain data and aligning trajectory-level entropy dynamics, matching full-shot performance with 32 target samples.
-
LLM-AutoDP: Automatic Data Processing via LLM Agents for Model Fine-tuning
LLM agents iteratively generate and optimize data processing strategies for fine-tuning, delivering over 80% win rates versus unprocessed data and 65% versus LLM-based AutoML baselines while cutting search time by up to 10x.
-
Training-Trajectory-Aware Token Selection
Training-Trajectory-Aware Token Selection (T3S) reconstructs the token-level training objective to overcome a performance bottleneck in continual distillation of reasoning capabilities from large to small language models.
-
Understanding and Improving Noisy Embedding Techniques in Instruction Finetuning
SymNoise applies symmetric noise to embeddings during instruction fine-tuning and reports 6.7% higher AlpacaEval scores than NEFTune on LLaMA-2-7B.
-
Cost-Aware Learning
Cost-Aware SGD samples by gradient-norm-to-cost ratio and is instantiated as Cost-Aware GRPO for length-dependent policy gradients, reducing tokens used in LLM RL while matching baseline accuracy.
-
OptProver: Bridging Olympiad and Optimization through Continual Training in Formal Theorem Proving
OptProver transfers formal theorem proving from Olympiad math to optimization via continual training, achieving SOTA Pass@1 and Pass@32 on a new Lean 4 benchmark while retaining general performance.
-
Rethinking Data Curation in LLM Training: Online Reweighting Offers Better Generalization than Offline Methods
ADAPT is an online reweighting framework for LLM training that outperforms offline data selection and mixing methods in cross-benchmark generalization under equal compute.
-
MM-LIMA: Less Is More for Alignment in Multi-Modal Datasets
MM-LIMA uses proposed quality metrics and a trainable selector to pick 200 high-quality multimodal instruction examples and outperforms MiniGPT-4 on evaluations.
-
Proxy Reward Internalization and Mechanistic Exploitation: A Learned Precursor to Reward Hacking and Its Generalization
Proxy RL produces a staged proxy-internalization capability that emerges before and predicts reward hacking in coding environments.
-
XekRung Technical Report
XekRung achieves state-of-the-art performance on cybersecurity benchmarks among same-scale models via tailored data synthesis and multi-stage training while retaining strong general capabilities.
-
Positional Encoding in Transformer-Based Time Series Models: A Survey
A survey of positional encoding methods in transformer-based time series models that evaluates fixed, learnable, relative, and hybrid approaches on classification tasks and links effectiveness to data characteristics.