PG-OT builds prompt-specific Pareto frontiers and applies distribution-aware optimal transport to improve multi-reward alignment while introducing JDR and JCR metrics to measure synergy and hacking.
Title resolution pending
29 Pith papers cite this work. Polarity classification is still indexing.
29
Pith papers citing it
representative citing papers
Post-Reasoning boosts LLM accuracy by reversing the usual answer-after-reasoning order, delivering mean relative gains of 17.37% across 117 model-benchmark pairs with zero extra cost.
ReflectMT internalizes reflection via two-stage RL to enable direct high-quality machine translation that outperforms explicit reasoning models like DeepSeek-R1 on WMT24 while using 94% fewer tokens.
Introduces the first large-scale 3D PET/CT dataset with fine-grained RoI annotations for Vietnamese and a graph-enhanced HiRRA framework that achieves SOTA report generation by modeling RoI dependencies.
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.
Short GRPO warm-up followed by offline DPO on informative rollouts matches or beats full GRPO on math reasoning benchmarks at substantially lower compute cost.
Weasel is a trajectory selection method that improves out-of-domain generalization for web agents while achieving 9.7-12.5x training speedups via importance-diversity optimization, AXTree pruning, and rationale style matching.
TOPOS creates high-fidelity 3D heads with fixed industry topology from single images via a specialized VAE with Perceiver Resampler and a rectified flow transformer.
OGLS-SD improves on-policy self-distillation stability and math reasoning performance by constructing an outcome-discriminative steering direction from contrasts between successful and failed teacher logits.
GTLM injects graph-aware attention biases into LLMs using only 0.015% extra parameters, enabling native graph processing that matches 7B models with a 1B model on text-attributed graph benchmarks.
MARLaaS enables concurrent RL fine-tuning across up to 32 tasks using LoRA adapters and a disaggregated asynchronous architecture, matching single-task performance while improving accelerator utilization by 4.3x and cutting end-to-end time by 85%.
RPSFT improves the in-domain versus out-of-domain performance trade-off during LLM supervised fine-tuning by penalizing rotations in pretrained singular subspaces as a proxy for loss-sensitive directions.
Current audio-language models fail to use clinical multimodal context for dysarthric speech recognition, but context-aware LoRA fine-tuning delivers large accuracy gains on the SAP dataset.
A training-free technique manipulates low-frequency noise in diffusion models to control image color and structure using low-frequency priors.
COMPASS uses semantic clustering on multilingual embeddings to select auxiliary data for PEFT adapters, outperforming linguistic-similarity baselines on multilingual benchmarks while supporting continual adaptation.
Math reasoning gains in LLMs rarely transfer to general domains; RL tuning generalizes while SFT causes forgetting and representation drift.
Zephyr-7B achieves state-of-the-art chat benchmark results among 7B models by distilling alignment via dDPO on AI feedback preferences, surpassing the 70B Llama-2-Chat model on MT-Bench with no human data required.
Orbax supplies a modular JAX-native checkpointing library that claims up to 3.5× faster saves and 2× faster loads than comparable PyTorch solutions.
Defines Entropy-Gradient Inversion as a negative entropy-gradient correlation fingerprinting LRM reasoning and proposes CorR-PO to embed it in RL regularization, claiming consistent outperformance on benchmarks.
Task-aware pruning improves OOD model performance by realigning distorted OOD layerwise norm and pairwise-distance profiles with the task-adapted geometry observed on ID inputs.
300 high-quality Stoic examples align small LLMs with inward virtues via preference optimization but leave outward cosmopolitan duties unlearned.
PRISM interleaves VLM perception and LLM reasoning via a dynamic goal-oriented question-answer pipeline to produce sharper scene descriptions, outperforming prior image-based models on ALFWorld and Room-to-Room.
GradShield is a data filtering technique using FIHS scores and adaptive thresholding to prevent safety misalignment in finetuned LLMs while preserving utility.
Qwen-Scope provides open-source sparse autoencoders for Qwen models that function as practical interfaces for steering, evaluating, data workflows, and optimizing large language models.
citing papers explorer
-
Post Reasoning: Improving the Performance of Non-Thinking Models at No Cost
Post-Reasoning boosts LLM accuracy by reversing the usual answer-after-reasoning order, delivering mean relative gains of 17.37% across 117 model-benchmark pairs with zero extra cost.
-
When Audio-Language Models Fail to Leverage Multimodal Context for Dysarthric Speech Recognition
Current audio-language models fail to use clinical multimodal context for dysarthric speech recognition, but context-aware LoRA fine-tuning delivers large accuracy gains on the SAP dataset.
-
Does Math Reasoning Improve General LLM Capabilities? Understanding Transferability of LLM Reasoning
Math reasoning gains in LLMs rarely transfer to general domains; RL tuning generalizes while SFT causes forgetting and representation drift.
-
Entropy-Gradient Inversion: Moving Toward Internal Mechanism of Large Reasoning Models
Defines Entropy-Gradient Inversion as a negative entropy-gradient correlation fingerprinting LRM reasoning and proposes CorR-PO to embed it in RL regularization, claiming consistent outperformance on benchmarks.
-
PRISM: Perception Reasoning Interleaved for Sequential Decision Making
PRISM interleaves VLM perception and LLM reasoning via a dynamic goal-oriented question-answer pipeline to produce sharper scene descriptions, outperforming prior image-based models on ALFWorld and Room-to-Room.