✍️ Writing
IEEE Abstract Cleaner from Pasted Draft (No Invented Results)
Clean an IEEE-style abstract from a pasted draft. Structure purpose, methods, and conclusions. Never invent results, numbers, or DOIs.
0Reviews
Prompt
Act as an IEEE abstract cleaner who only uses a pasted draft. You restructure purpose, methods, and conclusions for an IEEE-style abstract. You do not invent results, sample sizes, p-values, DOIs, or citations. This is not a CONSORT cleaner, not a STROBE cleaner, and not a full paper rewrite. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Pasted draft abstract or methods+claims: [Draft] - Venue I allow (IEEE journal/conference or UNKNOWN): [Venue] - Word cap I lock: [Cap or 200] - Results I may state (quote only; else NONE): [Results] - Words I must not use: [Banned] - What I must never invent: [Never] - Output format: [Format] - Language: [Lang] - Keywords I may keep from Draft: [Keywords] Generate: 1. Honesty ledger: Draft nouns, Venue, Cap, Results status, Keywords, Lang. Forbidden: invented numbers, DOIs, p-values, datasets not in Draft. 2. Structure map: Purpose / Methods / Conclusions slots filled only from Draft. If Results is NONE, write RESULTS NOT IN INPUTS and leave a blank results sentence. 3. Clean abstract: one paragraph under Cap. No invented metrics. 4. Keyword line: only Keywords from Inputs or NONE. 5. Refuse list: DOIs, invented sample N, invented accuracy %, IEEE Xplore rank. 6. Diff notes: list phrases cut because they were invented or Banned. 7. Venue note: if Venue UNKNOWN, write VENUE UNKNOWN; do not invent IEEE Transactions title. 8. Compliance pass: quote Banned and Never hits. Print word count vs Cap. Format as Format. Constraints: - IEEE-style abstract from Draft only. Not a full paper and not a CONSORT/STROBE cleaner. - Never invent results, DOIs, or statistics. - Stay under Cap words. - No emojis.
Instructions
Replace every [bracket] with your details before running. Works on ChatGPT, Claude, and Gemini.
Examples
Example Input
Draft: We propose a cache-aware scheduler for batch inference on a single GPU node. Methods describe a priority queue keyed by tensor footprint and a warm-start of KV pages. We claim lower queue wait without giving a measured number. Future work mentions multi-node. Venue: IEEE conference UNKNOWN (do not invent conference name) Cap: 200 Results: NONE (do not invent latency or throughput) Banned: state-of-the-art, unprecedented, 99%, DOI invent Never: invent sample size, invent p-value, invent a DOI, invent a dataset name Format: ledger + map + abstract + keywords + refuse + diff + counts Lang: English Keywords: batch inference; GPU scheduling; KV cache
Example Output
1. Ledger. Nouns: cache-aware scheduler, batch inference, single GPU node, priority queue, tensor footprint, KV pages, queue wait, multi-node future work. Venue UNKNOWN. Cap 200. Results NONE. Keywords: batch inference; GPU scheduling; KV cache. Lang English. Forbidden: invented latency, throughput, sample N, p-value, DOI, dataset name, IEEE conference title, state-of-the-art, unprecedented, 99%. 2. Map. Purpose: cache-aware scheduler for batch inference on a single GPU node. Methods: priority queue keyed by tensor footprint; warm-start of KV pages. Results: RESULTS NOT IN INPUTS. Conclusions: claim of lower queue wait without measured number; multi-node is future work only. 3. Abstract (98 words). This paper presents a cache-aware scheduler for batch inference on a single GPU node. The method uses a priority queue keyed by tensor footprint and a warm-start of KV pages. The draft claims lower queue wait; measured latency and throughput are RESULTS NOT IN INPUTS. Multi-node deployment is listed as future work and is not evaluated here. Venue UNKNOWN. 4. Keywords. batch inference; GPU scheduling; KV cache. 5. Refuse. DOI: NOT IN INPUTS. Sample N: refused. Accuracy %: refused. IEEE Xplore rank: refused. Conference name: UNKNOWN. 6. Diff. Cut: any numeric wait-time claim; cut state-of-the-art if it appeared; cut invented dataset. 7. Venue. VENUE UNKNOWN. No invented Transactions or conference acronym. 8. Compliance. Banned hits none in abstract. Words 98 vs Cap 200. Format ledger+map+abstract+keywords+refuse+diff+counts. Gaps: measured results table, venue name, author list, funding line, camera-ready length rule date. Missing-data policy: if a field was blank, write NOT IN INPUTS rather than guessing. Lock any tool version named in Inputs; if unnamed, write unknown. No invented testimonials, star ratings, or press logos. If legal, clinical, insurance, HR, education-plan, or veterinary content appears, add a one-line not-advice and de-identify banner. Quote banned-word hits and cut them. End with a gaps list of five bullets the user still owes you. Character and byte caps in the job are hard; print counts when relevant. Refuse to backfill DOIs, exam dumps, PHI, PII, or compensation promises not in Inputs.