🧠 AI - dungngo · Scour

Evaluating the Impact of Task Granularity on Catastrophic Forgetting in Continual Learning

⚙Systems Programming Academic

Optimisation over non-stationary distributions creates weirder minds

⚙Systems Programming

lesswrong.com·

Bridging the Morphology Gap: Adapting VLA Models to Dexterous Manipulation via Intent-Conditioned Fine-Tuning

🤖Robotics Academic

Not Just After One: Sleep-Inspired Replay Prevents Catastrophic Forgetting After Sequential Tasks

⚡Zig Academic

Preserving Plasticity in Continual Learning via Dynamical Isometry

🧠Neural Architecture Academic

Steering Multirobot Behavior via Closed-Loop Affine Activation Editing

🤖Robotics Academic

HydraCIL: Decoupled Class-Incremental Learning through Prototype-Guided Multi-Head Classifiers

📱Edge AI Academic

Listen, Look, and Learn: Learning Without Forgetting through SAM-Audio

🧠Neural Architecture Academic

Continual Quadruped Robots Coordination via Semantic Skill Discovery

🤖Robotics Academic

GUI-AC: Enhancing Continual Learning in GUI Agents

📱Edge AI Academic

Sparse Subspace-to-Expert Sharing for Task-Agnostic Continual Learning

🧠Neural Architecture Academic

TLRD: Teaching LLMs to Reason over Tabular Data with Tri-Level Rationale Distillation

💬NLP Academic

LargeMonitor: Monitoring Online Task-Free Continual Learning via Large Pretrained Models

🧠Neural Architecture Academic

Unsupervised Continual Clustering via Forward-Backward Knowledge Distillation

🧠Neural Architecture Academic

FOGO: Forgetting-aware Orthogonalization Optimizer

🧠Neural Architecture Academic

Two-Way Is Better Than One: Bidirectional Alignment with Cycle Consistency for Exemplar-Free Class-Incremental Learning

🧠Neural Architecture Academic

Two to Tango: Coupled Task-Reference Selection for Safe LLM Fine-tuning

⚙️LLVM Academic

The Fine-Tuning Trap: Evaluating Negative Transfer and the Role of PEFT in Sub-1B Mathematical Reasoning

🧠Neural Architecture Academic

Flow-DPPO: Divergence Proximal Policy Optimization for Flow Matching Models

🧠Neural Architecture Academic

Revisiting Prototype Rehearsal for Exemplar-Free Continual Learning: Manifold-Aware Boundary Sampling with Adaptive Class-Balanced Loss

🧠Neural Architecture Academic

Log in to enable infinite scrolling