AGENT SKILL

core-neurons

Build and debug snnTorch neuron, recurrent layer, state lifecycle, and batchnorm-through-time workflows.

License 许可
MIT
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
core-neurons
license
MIT
metadata
{ "disco-role": "operating" }
description
Build and debug snnTorch neuron, recurrent layer, state lifecycle, and batchnorm-through-time workflows.
disable-model-invocation
true

02 资源文件 8

03 SKILL.md 正文

core-neurons

Use this sub-skill when the task is to construct, combine, or debug snnTorch neuron and neuron-layer workflows.

Route here for

  • Core neuron construction and calls for Leaky, Synaptic, RLeaky, RSynaptic, Lapicque, Alpha, and DeltaLeaky.
  • Time-parallel or state-only neuron layers: LinearLeaky, StateLeaky, and LeakyParallel.
  • Recurrent and LSTM-like spiking layers: SConv2dLSTM, SLSTM, and AssociativeLeaky.
  • Hidden-state lifecycle choices: manual state tuples, init_hidden=True, reset_mem, class-level reset_hidden/detach_hidden, and snntorch.utils.reset.
  • Reset semantics, learnable decay/threshold flags, recurrent wiring choices, BatchNormTT1d/BatchNormTT2d, and GradedSpikes.

Do not handle here

  • Spike encodings, surrogate-gradient selection depth, loss functions, metrics, STDP, quantization recipes, or full training-loop design: route to encoding-training.
  • NIR export/import or NIR graph round trips: route to nir-interoperability.
  • Spike raster/count/trace/animation plotting: route to plotting.
  • Legacy neuromorphic dataset wrappers or dataset downloads: route to spikevision.

Operating workflow

  1. Identify whether the user needs a stepwise neuron cell, a time-major sequence layer, a recurrent cell, or a helper layer.
  2. Check the exact constructor, state tuple, and return contract in API reference.
  3. Pick the state-management pattern in workflows: manual states, init_hidden=True, stateless time-major layers, or LeakyParallel.
  4. Before changing model logic, triage state-count, reset, learnable-parameter, device, and shape failures with troubleshooting.
  5. For a safe local smoke check, run one of the bundled scripts below; they use synthetic tensors only and require no downloads.

Bundled scripts

Start with workflows for recipes and API reference for argument-level constraints.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351