AGENT SKILL

flow_recovery_diagnostics

Integrate reduced InterFlow velocity fields and summarize executable mechanism-faithful recovery diagnostics.

License 许可
Apache-2.0
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
flow_recovery_diagnostics
description
Integrate reduced InterFlow velocity fields and summarize executable mechanism-faithful recovery diagnostics.

02 资源文件 3

03 SKILL.md 正文

Flow Recovery Diagnostics

Use this skill when a recovery experiment must implement the stochastic-interpolant paper mechanism without reading the original repository. Do not use it to claim full CIFAR, ImageNet, or tabular reproduction unless the corresponding real datasets and training stack actually ran.

Inputs

  • Numeric base and target samples, time samples, and velocity parameters.
  • A declared full or reduced recovery target from module_plan.json.

Outputs

  • Deterministic numeric arrays, losses, gradients, or diagnostics depending on the module.
  • JSON-compatible evidence for recovery logs.

Workflow

  1. Build or consume stochastic-interpolant samples from endpoint pairs.
  2. Keep objective computation separate from ODE integration.
  3. Emit explicit checks that can be validated by recover-paper.

Validation

Run python -m pytest tests or validate the tree with validate_skill_tree.py --run-tests.

Limitations

The scripts are intentionally small and deterministic; they support bounded proxy recovery and are not a replacement for full neural image or tabular training.

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