proxy_fusion_recovery
Run a bounded proxy recovery that compares explicit/implicit-style fusion and performs a small optimizer step on fusion weights.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- proxy_fusion_recovery
- description
- Run a bounded proxy recovery that compares explicit/implicit-style fusion and performs a small optimizer step on fusion weights.
02 资源文件 2
- 脚本scripts/proxy_train.py3.1 KB
- 脚本tests/test_proxy_fusion_recovery.py448 B
03 SKILL.md 正文
Proxy Fusion Recovery
When To Use
Use this skill for bounded infrared-visible image-fusion analysis derived from arXiv:2206.09581, especially when preserving modality-specific features, comparing explicit/implicit mechanisms, or running proxy recovery.
Inputs
See the corresponding module document in the attempt directory. Inputs must be local text or numeric arrays; do not read any original paper repository.
Outputs
The skill returns deterministic JSON-compatible diagnostics suitable for downstream recovery and analysis.
Workflow
- Validate input shapes or text content.
- Execute the module-specific mechanism with deterministic defaults.
- Return compact evidence and numeric diagnostics.
- For recovery, keep proxy/reduced labels explicit.
Validation
Run python /share/project/yuyang/workspace/Paper2Skills/Paper2Skills-Agent/src/packages/paper2skills-agent/src/paper2skills/skills/module-to-skill/scripts/validate_skill_tree.py /share/project/yuyang/workspace/Paperbench/record/case13/extracted_skills_attempt_001/exorl_unsupervised_rl/proxy_fusion_recovery --run-tests.
Limitations
This skill implements a lightweight, reusable proxy for the paper mechanism, not the full 21-test-set benchmark from the paper.
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).