ln_summary_protocol
Build likelihood-free linear-nonlinear neural simulations and STA/firing-rate summaries for SNPE-style posterior recovery.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- ln_summary_protocol
- description
- Build likelihood-free linear-nonlinear neural simulations and STA/firing-rate summaries for SNPE-style posterior recovery.
02 资源文件 2
- 脚本scripts/ln_summary.py2.1 KB
- 脚本tests/test_ln_summary.py796 B
03 SKILL.md 正文
LN Summary Protocol
Use this skill when a recovery or experiment needs a small mechanistic neural simulator with summary features matching the paper's LN receptive-field example. Do not use it as a generic classifier or as a likelihood evaluator.
Inputs
theta: receptive-field/filter vector.n_stimuli,dimension, andseedfor deterministic stimulus generation.- Optional
biasandgainfor Bernoulli spike probabilities.
Outputs
- A JSON-compatible item containing
theta,stimuli,spikes,summary,sta,firing_rate, and metadata.
Workflow
- Generate white-noise stimuli from a seeded RNG.
- Apply a logistic nonlinearity to stimulus-filter projections.
- Sample Bernoulli spikes without evaluating a likelihood.
- Compute STA from spike-triggering stimuli and append firing rate.
- Save the generated data item when used in recovery.
Validation
Run python tests/test_ln_summary.py from this skill directory.
Limitations
This is a reduced LN simulator for fast mechanism checks. It is not the full paper implementation and does not reproduce high-dimensional CNN/MDN experiments.
04 同仓库更多 Skills
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.
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.
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.
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.
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.
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).