representation_distance_scoring
Compute class-center representation distances for Moderate-DS style data selection from labeled feature records.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- representation_distance_scoring
- description
- Compute class-center representation distances for Moderate-DS style data selection from labeled feature records.
02 资源文件 2
- 脚本scripts/score_representations.py2.4 KB
- 脚本tests/test_scoring.py705 B
03 SKILL.md 正文
Representation Distance Scoring
Use this skill when a recovery or application has labeled examples with numeric hidden representations and needs the Moderate-DS distance score. Do not use it to train a model, choose the coreset size, or evaluate downstream accuracy; those are separate modules.
Inputs
- A JSON list of records with
id,label, andrepresentationfields. - Numeric representation vectors must have consistent dimensionality.
Outputs
- Class centers keyed by label.
- Per-record Euclidean distance scores from each representation to its own class center.
Workflow
- Validate record ids, labels, and representation dimensions.
- Average representation vectors separately for every class label.
- Compute the Euclidean norm between each record representation and the matching class center.
- Return stable diagnostics that preserve the input ids and labels.
Validation
Run python scripts/score_representations.py --self-test or python -m pytest tests when pytest is available. The deterministic tests cover centers, distances, and schema failures.
Limitations
This skill assumes representations already exist. It intentionally does not read the original paper repository, access model internals, or perform data augmentation.
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).