Voyager achieves superior lifelong learning in Minecraft by combining an automatic exploration curriculum, a library of executable skills, and iterative LLM prompting with environment feedback, yielding 3.3x more unique items and 15.3x faster milestone unlocks than prior methods while generalizing技能
hub
A comprehensive survey of continual learning: Theory, method and application
15 Pith papers cite this work, alongside 57 external citations. Polarity classification is still indexing.
15
Pith papers citing it
57
external citations · external index
hub tools
citation-role summary
background 4
citation-polarity summary
roles
background 4representative citing papers
FBCC introduces unsupervised continual clustering via a teacher-student forward-backward distillation process that outperforms baselines on clustering accuracy while reducing catastrophic forgetting.
TFGN is an architectural overlay for transformers enabling task-free, replay-free continual pre-training across heterogeneous domains at LLM scale with near-zero backward transfer and high gradient orthogonality.
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.
Non-model gains via inference, systems, and assets can drive AI capabilities independently of base models, requiring governance beyond model-level evaluation and mitigation.
DryRUN lets LLMs create their own test inputs and run internal simulations for self-correcting code generation, matching the performance of test-dependent methods like CodeSIM on LiveCodeBench without public tests or external signals.
EvolveR enables LLM agents to self-evolve via a closed loop of distilling interaction trajectories into strategic principles offline and retrieving them to guide online decisions with policy reinforcement, yielding better results on multi-hop QA benchmarks.
A plug-and-play KL regularizer that masks the target token and renormalizes probabilities to improve the learning-forgetting trade-off in LoRA adaptation of LLMs.
SF-NorMuon is a new schedule-free spectral optimizer that closes the gap with tuned AdamW on 125M-772M parameter models across 1-8x Chinchilla horizons while providing stationarity guarantees.
HEDP uses energy regularization inspired by Helmholtz free energy plus hybrid energy-distance weighting in prompts to improve domain selection and achieve a 2.57% accuracy gain on benchmarks like CORe50 while mitigating catastrophic forgetting.
A kernel plasticity approach in Hebbian DNNs for incremental sound classification achieves 76.3% accuracy over five steps on ESC-50, outperforming the 68.7% baseline without plasticity.
RBF is a hybrid edge-HPC architecture that decouples low-latency edge inference using surrogate models from asynchronous HPC-driven model updates for simulation-bounded cyber-physical systems.
LIFE is a proposed agentic framework that combines four components to enable incremental, flexible, and energy-efficient continual learning for HPC operations such as latency spike mitigation.
The paper surveys the origins, frameworks, applications, and open challenges of AI agents built on large language models.
Domain-incremental audio classification via frozen domain-specific experts plus prototype classifier on concatenated features yields 78.15% micro / 77.03% macro accuracy on the DCASE 2026 Task 7 development set.
citing papers explorer
-
Voyager: An Open-Ended Embodied Agent with Large Language Models
Voyager achieves superior lifelong learning in Minecraft by combining an automatic exploration curriculum, a library of executable skills, and iterative LLM prompting with environment feedback, yielding 3.3x more unique items and 15.3x faster milestone unlocks than prior methods while generalizing技能
-
Unsupervised Continual Clustering via Forward-Backward Knowledge Distillation
FBCC introduces unsupervised continual clustering via a teacher-student forward-backward distillation process that outperforms baselines on clustering accuracy while reducing catastrophic forgetting.
-
TFGN: Task-Free, Replay-Free Continual Pre-Training Without Catastrophic Forgetting at LLM Scale
TFGN is an architectural overlay for transformers enabling task-free, replay-free continual pre-training across heterogeneous domains at LLM scale with near-zero backward transfer and high gradient orthogonality.
-
Rotation-Preserving Supervised Fine-Tuning
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.
-
Comprehensive AI governance requires addressing non-model gains
Non-model gains via inference, systems, and assets can drive AI capabilities independently of base models, requiring governance beyond model-level evaluation and mitigation.
-
You Don't Need Public Tests to Generate Correct Code
DryRUN lets LLMs create their own test inputs and run internal simulations for self-correcting code generation, matching the performance of test-dependent methods like CodeSIM on LiveCodeBench without public tests or external signals.
-
EvolveR: Self-Evolving LLM Agents through an Experience-Driven Lifecycle
EvolveR enables LLM agents to self-evolve via a closed loop of distilling interaction trajectories into strategic principles offline and retrieving them to guide online decisions with policy reinforcement, yielding better results on multi-hop QA benchmarks.
-
Mask the Target: A Plug-and-Play Regularizer Against LoRA Forgetting
A plug-and-play KL regularizer that masks the target token and renormalizes probabilities to improve the learning-forgetting trade-off in LoRA adaptation of LLMs.
-
Anytime Training with Schedule-Free Spectral Optimization
SF-NorMuon is a new schedule-free spectral optimizer that closes the gap with tuned AdamW on 125M-772M parameter models across 1-8x Chinchilla horizons while providing stationarity guarantees.
-
HEDP: A Hybrid Energy-Distance Prompt-based Framework for Domain Incremental Learning
HEDP uses energy regularization inspired by Helmholtz free energy plus hybrid energy-distance weighting in prompts to improve domain selection and achieve a 2.57% accuracy gain on benchmarks like CORe50 while mitigating catastrophic forgetting.
-
Incremental learning for audio classification with Hebbian Deep Neural Networks
A kernel plasticity approach in Hebbian DNNs for incremental sound classification achieves 76.3% accuracy over five steps on ESC-50, outperforming the 68.7% baseline without plasticity.
-
Hybrid Edge-HPC Systems for Low-Latency Data-Driven Inference
RBF is a hybrid edge-HPC architecture that decouples low-latency edge inference using surrogate models from asynchronous HPC-driven model updates for simulation-bounded cyber-physical systems.
-
LIFE -- an energy efficient advanced continual learning agentic AI framework for frontier systems
LIFE is a proposed agentic framework that combines four components to enable incremental, flexible, and energy-efficient continual learning for HPC operations such as latency spike mitigation.
-
The Rise and Potential of Large Language Model Based Agents: A Survey
The paper surveys the origins, frameworks, applications, and open challenges of AI agents built on large language models.
-
Domain-incremental audio classification using domain-specific experts and prototype classifier
Domain-incremental audio classification via frozen domain-specific experts plus prototype classifier on concatenated features yields 78.15% micro / 77.03% macro accuracy on the DCASE 2026 Task 7 development set.