AIBullisharXiv – CS AI · 3d ago7/10
🧠Researchers demonstrate that Evolution Strategies (ES) can effectively fine-tune large language models without catastrophic forgetting of prior tasks, contrary to recent concerns. By introducing Anchored Weight Decay (AWD), a regularization technique that constrains optimization toward initial parameters, the work shows ES-based continual learning is viable and computationally efficient compared to reinforcement learning approaches.
AIBullisharXiv – CS AI · May 97/10
🧠Researchers present a statistical-physics framework explaining how large language models develop multi-step reasoning through reinforcement learning with verifiable rewards (RLVR), modeling the process as inverse tree freezing in a concept network. They propose Annealed-RLVR, a timing-optimized training method that outperforms standard RLVR by applying supervised fine-tuning at peak frustration rather than after convergence, preventing policy collapse.
AIBullisharXiv – CS AI · May 77/10
🧠Researchers propose Anchored Learning, a new fine-tuning method that prevents catastrophic forgetting in large language models by controlling distributional drift through a dynamically evolving reference anchor. The technique achieves near-optimal performance gains while reducing degradation from over 53% to under 5% on benchmark tasks.
AIBullisharXiv – CS AI · May 77/10
🧠Researchers propose a novel framework that models language model memory as a Markov transition matrix, enabling efficient incorporation of new knowledge without catastrophic forgetting. The approach requires only linear sample complexity in the number of existing tokens and achieves zero forgetting through minimal parameter updates via an embedding-tuning algorithm.
AIBullisharXiv – CS AI · May 77/10
🧠Researchers propose skill neologisms—soft tokens added to LLM vocabularies—as a scalable approach to continual learning that enables models to acquire new capabilities without catastrophic forgetting or weight updates. The method demonstrates that independently trained skill tokens can compose zero-shot and work with out-of-distribution tasks, offering a practical alternative to fine-tuning.
AIBullisharXiv – CS AI · Apr 147/10
🧠Researchers introduce soul.py, an open-source architecture addressing catastrophic forgetting in AI agents by distributing identity across multiple memory systems rather than centralizing it. The framework implements persistent identity through separable components and a hybrid RAG+RLM retrieval system, drawing inspiration from how human memory survives neurological damage.
AIBullisharXiv – CS AI · Apr 147/10
🧠Researchers propose VaCoAl, a hyperdimensional computing architecture that combines sparse distributed memory with Galois-field algebra to address limitations in modern AI systems like catastrophic forgetting and the binding problem. The deterministic system demonstrates emergent properties equivalent to spike-timing-dependent plasticity and achieves multi-hop reasoning across 25.5M paths in knowledge graphs, positioning it as a complementary third paradigm to large language models.
AINeutralarXiv – CS AI · Apr 107/10
🧠Researchers introduce the Informational Buildup Framework (IBF), a new approach to continual learning that eliminates catastrophic forgetting by treating information as structural alignment rather than stored parameters. The framework demonstrates superior performance across multiple domains including chess and image classification, achieving near-zero forgetting without requiring raw data replay.
AIBullisharXiv – CS AI · Mar 177/10
🧠Researchers introduce SCAN, a new framework for editing Large Language Models that prevents catastrophic forgetting during sequential knowledge updates. The method uses sparse circuit manipulation instead of dense parameter changes, maintaining model performance even after 3,000 sequential edits across major models like Gemma2, Qwen3, and Llama3.1.
🧠 Llama
AIBullisharXiv – CS AI · Mar 56/10
🧠Researchers discovered that pretrained Vision-Language-Action (VLA) models demonstrate remarkable resistance to catastrophic forgetting in continual learning scenarios, unlike smaller models trained from scratch. Simple Experience Replay techniques achieve near-zero forgetting with minimal replay data, suggesting large-scale pretraining fundamentally changes continual learning dynamics for robotics applications.
AIBullisharXiv – CS AI · Mar 47/103
🧠Researchers have identified a critical flaw in reinforcement learning fine-tuning of large language models that causes degradation in multi-attempt performance despite improvements in single attempts. Their proposed solution, Diversity-Preserving Hybrid RL (DPH-RL), uses mass-covering f-divergences to maintain model diversity and prevent catastrophic forgetting while improving training efficiency.
AIBullisharXiv – CS AI · Mar 46/103
🧠Researchers developed cPNN (Continuous Progressive Neural Networks), a new AI architecture that handles evolving data streams with temporal dependencies while avoiding catastrophic forgetting. The system addresses concept drift in time series data by combining recurrent neural networks with progressive learning techniques, showing quick adaptation to new concepts.
AIBullisharXiv – CS AI · Mar 37/103
🧠Researchers introduce Dream2Learn (D2L), a continual learning framework that enables AI models to generate synthetic training data from their own internal representations, mimicking human dreaming for knowledge consolidation. The system creates novel 'dreamed classes' using diffusion models to improve forward knowledge transfer and prevent catastrophic forgetting in neural networks.
AIBullisharXiv – CS AI · Feb 277/106
🧠Researchers introduce GraftLLM, a new method for transferring knowledge between large language models using 'SkillPack' format that preserves capabilities while avoiding catastrophic forgetting. The approach enables efficient model fusion and continual learning for heterogeneous models through modular knowledge storage.
AIBullisharXiv – CS AI · 3d ago6/10
🧠Researchers introduce TRACER, a novel finetuning method for multimodal AI models that addresses catastrophic forgetting and out-of-distribution robustness degradation. By replacing standard Exponential Moving Average teachers with Weighted Moving Average teachers and combining contrastive learning with multi-perspective distillation, the approach demonstrates consistent performance gains across CLIP backbone architectures without hyperparameter sensitivity.
AINeutralarXiv – CS AI · 3d ago6/10
🧠Researchers demonstrate that reinforcement learning (RL) preserves internal computational circuits in large language models better than supervised fine-tuning (SFT) during task adaptation. Using a new metric called differential circuit vulnerability on Qwen2.5-3B-Instruct, they reveal a mechanistic trade-off: SFT adapts faster but causes substantial circuit disruption and capability forgetting, while RL maintains base model circuits at the cost of slower learning.
AINeutralarXiv – CS AI · 4d ago6/10
🧠Researchers introduce SAME, a new approach for training Multimodal Large Language Models that can continuously learn new tasks without forgetting previous capabilities. The method addresses fundamental problems in continual learning by stabilizing how AI systems route tasks to specialized expert networks and preventing knowledge degradation over time.
AINeutralarXiv – CS AI · 5d ago6/10
🧠Researchers introduce a counterfactual-free circuit discovery method adapted for unstructured natural text, enabling Circuit-Targeted Supervised Fine-Tuning (CT-SFT) that improves low-resource model adaptation while preserving performance on source tasks and preventing catastrophic forgetting.
AINeutralarXiv – CS AI · May 126/10
🧠Researchers introduce UFO, a framework addressing robust continual graph learning by simultaneously tackling catastrophic forgetting and noisy data supervision in evolving graphs. The method uses flow-based generative modeling to mitigate forgetting and instance-level reliability scoring to handle corrupted labels, demonstrating superior performance across benchmark datasets.
AINeutralarXiv – CS AI · May 96/10
🧠Researchers demonstrate that using the same optimizer during both pretraining and finetuning of large language models reduces catastrophic forgetting while maintaining or improving task performance. This "optimizer-model consistency" effect suggests optimizers create regularization patterns that preserve learned knowledge, with implications for efficient model adaptation strategies.
AINeutralarXiv – CS AI · May 96/10
🧠Researchers propose DREE, a novel lifelong learning framework for neural vehicle routing problem solvers that handles continually drifting task patterns with limited training resources per task. The approach addresses a gap in existing methods by managing catastrophic forgetting while learning sequential tasks in real-world logistics scenarios where problem patterns shift over time.
AINeutralarXiv – CS AI · May 96/10
🧠Researchers introduce CRAFT, a continual learning framework for large language models that prevents catastrophic forgetting by learning low-rank interventions on hidden representations rather than updating model weights. The three-stage approach uses KL divergence-based routing and merging to enable models to acquire new capabilities while maintaining performance on previously learned tasks.
AINeutralarXiv – CS AI · May 96/10
🧠Researchers introduce HEDP, a domain incremental learning framework that enables AI models to adapt to new data domains without retraining by combining energy-based regularization with distance-based weighting mechanisms. The approach demonstrates a 2.57% accuracy improvement on unseen domains while reducing catastrophic forgetting, addressing a critical challenge in continuous learning systems.
AIBullisharXiv – CS AI · Apr 206/10
🧠Researchers introduce JumpLoRA, a novel framework that uses sparse adapters with JumpReLU gating to enable continual learning in large language models while mitigating catastrophic forgetting. The method dynamically isolates parameters across tasks, outperforming existing state-of-the-art approaches like ELLA and significantly improving IncLoRA performance.
AIBullisharXiv – CS AI · Apr 156/10
🧠Researchers propose Joint Flashback Adaptation, a novel method to address catastrophic forgetting in large language models during incremental task learning. The approach uses limited prompts from previous tasks combined with latent task interpolation, demonstrating improved performance across 1000+ instruction-following and reasoning tasks without requiring full replay data.
