Active lifting of data distributions via latent-sequence sampling and max-rate uncertainty reduction formally derives slow-thinking LLMs and places them on representation and sampler hierarchies that can be climbed.
Ravr: Reference-answer-guided variational reasoning for large language models
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 5roles
background 1polarities
background 1representative citing papers
Variable-length autoregressive latent sequences, trained as variational inference with a PPO-style objective, give robot policies adaptive test-time compute and yield a reusable action tokenizer.
This survey introduces the Generate-Filter-Control-Replay (GFCR) taxonomy to structure rollout pipelines for RL-based post-training of reasoning LLMs.
AMVL applies bidirectional KL calibration to align answer-agnostic prior with answer-conditioned posterior in variational multimodal reasoning, reducing leakage and yielding +10.83 average gain on BLINK benchmark.
VPG-EA applies variational posterior guidance and efficiency-aware distillation to compress LLM reasoning chains while preserving performance.
citing papers explorer
-
A First-Principles Theory of Slow Thinking and Active Perception
Active lifting of data distributions via latent-sequence sampling and max-rate uncertainty reduction formally derives slow-thinking LLMs and places them on representation and sampler hierarchies that can be climbed.
-
Latent Memory Palace: Reasoning for Control as Autoregressive Variational Inference
Variable-length autoregressive latent sequences, trained as variational inference with a PPO-style objective, give robot policies adaptive test-time compute and yield a reusable action tokenizer.
-
Generate, Filter, Control, Replay: A Comprehensive Survey of Rollout Strategies for LLM Reinforcement Learning
This survey introduces the Generate-Filter-Control-Replay (GFCR) taxonomy to structure rollout pipelines for RL-based post-training of reasoning LLMs.
-
Multimodal Continuous Reasoning via Asymmetric Mutual Variational Learning
AMVL applies bidirectional KL calibration to align answer-agnostic prior with answer-conditioned posterior in variational multimodal reasoning, reducing leakage and yielding +10.83 average gain on BLINK benchmark.
-
Efficient LLM Reasoning via Variational Posterior Guidance with Efficiency Awareness
VPG-EA applies variational posterior guidance and efficiency-aware distillation to compress LLM reasoning chains while preserving performance.