MCPcopy Create free account
hub / github.com/917Dhj/DeepPaperNote / extract_mechanism_flow_sentences

Function extract_mechanism_flow_sentences

scripts/common.py:2336–2418  ·  view source on GitHub ↗
(text: str, *, limit: int = 8)

Source from the content-addressed store, hash-verified

2334
2335
2336def extract_mechanism_flow_sentences(text: str, *, limit: int = 8) -> list[str]:
2337 claims: list[str] = []
2338 seen = set()
2339 action_tokens = [
2340 "encode",
2341 "encoded",
2342 "encoding",
2343 "extract",
2344 "extracted",
2345 "project",
2346 "projected",
2347 "pool",
2348 "pooled",
2349 "fuse",
2350 "fused",
2351 "fusion",
2352 "concat",
2353 "concatenate",
2354 "query",
2355 "queried",
2356 "align",
2357 "aligned",
2358 "compress",
2359 "compressed",
2360 "send",
2361 "sent",
2362 "feed",
2363 "fed",
2364 "generate",
2365 "generated",
2366 "predict",
2367 "predicted",
2368 "decode",
2369 "decoded",
2370 "update",
2371 "updated",
2372 "freeze",
2373 "frozen",
2374 "fine-tune",
2375 "finetune",
2376 ]
2377 flow_tokens = [
2378 "input",
2379 "inputs",
2380 "output",
2381 "outputs",
2382 "token",
2383 "tokens",
2384 "feature",
2385 "features",
2386 "representation",
2387 "representations",
2388 "encoder",
2389 "decoder",
2390 "attention",
2391 "module",
2392 "modules",
2393 "llm",

Calls 3

split_sentencesFunction · 0.85
normalize_whitespaceFunction · 0.85
normalize_titleFunction · 0.85