RaPO reduces catastrophic forgetting in visual continual learning by shaping rewards around policy drift and stabilizing advantages with cross-task exponential moving averages during reinforcement fine-tuning of multimodal models.
Continual learning of large language models: A comprehensive survey.ACM Computing Surveys, 58(5):1–42
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 4
citation-polarity summary
years
2026 8verdicts
UNVERDICTED 8roles
background 4polarities
background 4representative citing papers
TRC² is a brain-inspired decoder-only architecture that localizes fast plasticity and uses thalamic and hippocampal pathways to substantially reduce cumulative forgetting in sequential language model training on streams like C4, WikiText-103, and GSM8K.
MedEvoEval is an executable longitudinal evaluation framework that converts medical cases into action-gated simulated episodes to track how doctor agents evolve decision-making, resource use, and experience across multiple encounters.
MAP improves LLM agent reasoning by constructing a structured cognitive map of the environment before task execution, yielding performance gains on benchmarks like ARC-AGI-3 and superior training data via the new MAP-2K dataset.
Forgetting in LLM continual post-training is a geometry conflict between task-induced covariance structures and the evolving model state, controlled by gating Wasserstein barycenter merging on measured conflict.
SCoL trains LLMs via meta-reinforcement learning to generate layer-specific update instructions that improve knowledge acquisition and retention from context streams over standard baselines.
ConSFT is a gradient-scaling fine-tuning objective for flow-matching VLAs that bounds parameter disruption via model-confidence weighting, yielding over 20% better capability retention than vanilla SFT on LIBERO and RoboTwin.
In the Flux environment, RL agents with explicit latent state access achieve ~79% win rate versus ~11% for LLMs on long-horizon tasks, illustrating limitations of sequence prediction for dynamic reasoning.
citing papers explorer
-
Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning
RaPO reduces catastrophic forgetting in visual continual learning by shaping rewards around policy drift and stabilizing advantages with cross-task exponential moving averages during reinforcement fine-tuning of multimodal models.
-
Efficient Continual Learning in Language Models via Thalamically Routed Cortical Columns
TRC² is a brain-inspired decoder-only architecture that localizes fast plasticity and uses thalamic and hippocampal pathways to substantially reduce cumulative forgetting in sequential language model training on streams like C4, WikiText-103, and GSM8K.
-
MedEvoEval: Evaluating Continual Evolution of Doctor Agents through Simulated Clinical Episodes
MedEvoEval is an executable longitudinal evaluation framework that converts medical cases into action-gated simulated episodes to track how doctor agents evolve decision-making, resource use, and experience across multiple encounters.
-
MAP: A Map-then-Act Paradigm for Long-Horizon Interactive Agent Reasoning
MAP improves LLM agent reasoning by constructing a structured cognitive map of the environment before task execution, yielding performance gains on benchmarks like ARC-AGI-3 and superior training data via the new MAP-2K dataset.
-
Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training
Forgetting in LLM continual post-training is a geometry conflict between task-induced covariance structures and the evolving model state, controlled by gating Wasserstein barycenter merging on measured conflict.
-
Self-Consolidating Language Models: Continual Knowledge Incorporation from Context
SCoL trains LLMs via meta-reinforcement learning to generate layer-specific update instructions that improve knowledge acquisition and retention from context streams over standard baselines.
-
Preserving Foundational Capabilities in Flow-Matching VLAs through Conservative SFT
ConSFT is a gradient-scaling fine-tuning objective for flow-matching VLAs that bounds parameter disruption via model-confidence weighting, yielding over 20% better capability retention than vanilla SFT on LIBERO and RoboTwin.
-
Why We Need World Models for AGI: Where LLMs Fail and How World Models May Outperform
In the Flux environment, RL agents with explicit latent state access achieve ~79% win rate versus ~11% for LLMs on long-horizon tasks, illustrating limitations of sequence prediction for dynamic reasoning.