Saliency-driven Experience Replay for Continual Learning (NeurIPS2024)
-
Updated
Mar 14, 2025 - Python
#
selective-attention
Here are 5 public repositories matching this topic...
Qwen3.8-27B at 256k context on a 16 GB AMD RX 9070 XT (llama.cpp Vulkan): measured, quality-gated, reproducible. 56.6-58.5 t/s at working depth, ~42-46 t/s at 200k resident with a selective-attention tile-skip.
benchmark amd gpu vulkan quantization selective-attention long-context llama-cpp llm-inference qwen gguf rx-9070-xt
-
Updated
Aug 24, 2026 - Shell
Openai-gym like environment introducing Selective Attention problem
-
Updated
Apr 27, 2019 - Python
Synthetic Thalamus is a biologically-inspired neural attention mechanism that mimics the brain's thalamic circuits. This project implements a selective information bottleneck with phase-based synchronization, contrastive learning, and recurrent feedback loops.
neural-attention thalamus information-bottleneck phase-synchronization selective-attention contrastive-learning transformer-enhancement
-
Updated
Apr 16, 2025 - Python
I built a benchmark to evaluate the Attention faculty in isolation, measuring how architectures maintain focus, filter noise, and progress toward AGI.
-
Updated
Sep 27, 2026 - Python
Add this topic to your repo
To associate your repository with the selective-attention topic, visit your repo's landing page and select "manage topics."