Showing results for "passage-ranking"
Design a plan to self-improve ranking/retrieval safely: offline eval sets, interleaving/AB tests, bias checks, and rollback on metric drops.
Author: Assistant
Category: safe-self-improving-ai | Model: gpt-5.2
Design snippet generation: pick supporting passages, highlight query terms/entities, and store offsets for UI rendering. Include how to avoid misleading snippets.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Create a source scoring rubric: domain trust, author expertise, citations, recency, and conflicts of interest. Output a weighted scoring model and how to tune it per domain.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Design a URL frontier with ranking signals: relevance to query, freshness, link context, domain trust, and duplication risk. Include pseudocode-level description and metrics.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Design a ranking formula: lexical relevance, semantic similarity, trust score, freshness decay, and diversity penalty. Include how to tune weights and validate offline/online.
Author: Assistant
Category: research-bot | Model: GPT-5.2
Design a ranked climb strategy: warm-up, session limits, tilt control, hero/role pool strategy, and performance metrics beyond win/loss. Include a ‘bad day’ protocol to stop bleeding rating.
Author: Assistant
Category: gaming-metaverse | Model: gpt-5.2
Build a 30-day academic vocabulary system: word families, collocations, spaced repetition schedule, and a method to learn from reading passages. Include a template for a personal vocab deck.
Author: Assistant
Category: toefl-2026 | Model: gpt-4o
Create a memory architecture: scratchpad, episodic memory, semantic memory, and “facts of record.” Include retention rules, privacy, deduplication, and retrieval ranking.
Author: Assistant
Category: agent-architecture | Model: GPT-5.2
Train a simple LTR model with handcrafted features (BM25 score, title match, click priors). Provide cross-validation setup and feature importance chart.
Author: Assistant
Category: ml-lab-IR | Model: gpt-4o
Run a bias assessment on search ranking: define fairness metrics, measure on subsets, and propose mitigations and business impact.
Author: Assistant
Category: ethics-and-impact-IR-MBA | Model: gpt-4o
Answer a design prompt: scholarly search with citation grounding. Provide indexing pipeline, re-ranking, and dedupe heuristics with diagrams.
Author: Assistant
Category: advanced-design-exercise-IR | Model: gpt-4o
Architect a retrieval stack with hybrid search, temporal decay, dedup, and passage-level citation anchors. Define fact-grounding checks and failure messages; include freshness reindex cadence.
Author: Assistant
Category: retrieval-grounding-LLM | Model: gpt-4o
Recommend when to use LoRA/QLoRA vs full finetune. Define rank search, target layers, and quantization-aware adapters. Include memory/perf tables per GPU class.
Author: Assistant
Category: parameter-efficient-tuning-LLM | Model: gpt-4o
Create a min-max policy for critical spares in Belo Horizonte. ABC rank, lead time buffers, consignment options, and stock-out risk modeling.
Author: Assistant
Category: MRO-inventory-optimization | Model: gpt-4o
Act as an aerospace systems lead. Draft a concept of operations (CONOPS) for a 4–6 seat eVTOL. Specify mission profile, range, payload, speed, reserves, and turnaround times. Produce a trade study com...
Author: Assistant
Category: aerospace-design | Model: gpt-4o
Design a time-decay model that boosts fresh, authoritative sources. Include publisher priors, sitemap/news bias, and re-crawl triggers. Output evaluation with time-split test sets.
Author: Assistant
Category: ranking-freshness | Model: gpt-4o
Design a UI: inline citations, hover previews, source quality badges, and ‘expand to passage’. Include keyboard flows and accessibility specs (WCAG). Output wireframes and event schema.
Author: Assistant
Category: product-UX-research | Model: gpt-4o
Specify a KG-augmented RAG: entity/link extraction → KG lookup → vector recall → late fusion ranking. Provide ranking features (entity overlap, path length), and a sample Gremlin/SPARQL query pack.
Author: Assistant
Category: hybrid-retrieval-kg | Model: gpt-4o
Build an eval that scores ground-truth attribution (exact passage match), answer faithfulness, and coverage. Provide dataset schema and a nightly regression plan.
Author: Assistant
Category: evaluation-frameworks | Model: gpt-4o
Train an LTR model to order passages by usefulness. Define features (BM25, dense score, novelty, redundancy), labels, and offline/online eval plan.
Author: Assistant
Category: retrieval-ranking-ml | Model: gpt-4o
Provide a mixed set (email, notice, article, multi-passage) with questions for gist/detail/inference. Add timing splits and skimming checkboxes.
Author: Assistant
Category: reading-sets-TOEIC | Model: gpt-4o
Create a TOEFL-style reading passage (700–900 words) with 10 questions covering gist, detail, inference, function, and vocab-in-context. Include a skim/scan stopwatch drill, answer key, and why-wrong ...
Author: Assistant
Category: reading | Model: gpt-4o
Create a power integrity plan using backside power delivery and buried power rails. Include PDN topology, stripe pitch, via stacking, decap budgeting, vectorless vs vector-based IR analysis, EM rules,...
Author: Assistant
Category: chip-design | Model: gpt-4
Plan a winter street‑food crawl in China around northern dumplings & steamed buns. Output a table with: rank, stall/name, exact spot (market/station), signature dish, spice/heat index (0–5), best time...
Author: Inspire Search Corp.
Category: Culinary (Winter) | Model: gpt-5-thinking
Plan a winter street‑food crawl in Japan around regional nabe (yose, chanko, motsu). Output a table with: rank, stall/name, exact spot (market/station), signature dish, spice/heat index (0–5), best ti...
Author: Inspire Search Corp.
Category: Culinary (Winter) | Model: gpt-5-thinking
Create a symmetry/leading‑lines scavenger list for Manila, Philippines. Provide 15 locations where converging lines or reflections dominate. Deliver as a table: rank, spot, why_it_works, recommended_f...
Author: Inspire Search Corp.
Category: Travel Photography (Asia) | Model: gpt-5-thinking
Goal: generate CRO hypotheses backed by evidence. Data: funnels, heatmaps, surveys, NPS, session replays. Steps: 1) Aggregate pain points; 2) Map to heuristics (clarity, friction, motivation); 3) Prio...
Author: Tsubasa Kato
Category: Web Analytics | Model: GPT-5 Thinking
Goal: audit source/medium quality and UTM hygiene for {START_DATE}–{END_DATE}. Data: GA4 BigQuery export `{PROJECT}.{DATASET}.events_*`, ad platform logs, server-side UTMs. Steps: 1) Classify sessions...
Author: Tsubasa Kato
Category: Web Analytics | Model: GPT-5 Thinking
Use daily global headlines to produce a ranked list of emerging risks by impact and likelihood. Provide early warning for possible black-swan developments.
Author: Tsubasa Kato
Category: Risk Intelligence | Model: GPT-5
You are a talent scout. Given {niches} and {regions}, build a ranked long‑list of emerging creators (1k–50k followers) with high up‑and‑to‑the‑right momentum. Method: use public signals (posting caden...
Author: Tsubasa Kato
Category: Market Research | Model: gpt-5
Create a market map for {space}. Buckets: Infra, Tooling, Apps, Services, Community. Propose a 0‑100 scoring rubric with weights for Momentum, Distribution, Moat, Team, Signal Quality. Score {entities...
Author: Tsubasa Kato
Category: Market Research | Model: gpt-5
Generate 10 diverse ideas for {{topic}} across people, process, product, promotion, and platform. Score each 1–5 for impact, effort, fun. Present the ranked top 3 with concrete next steps and a 24-hou...
Author: ChatGPT
Category: Brainstorming | Model: gpt-5
For {{model_family}} training on {{hardware}}: - Estimate energy/emissions baseline - Apply: low-rank adapters, distillation, AMP, token pruning, data curation - Carbon-aware job placement Provide a t...
Author: Tsubasa Kato
Category: mlops | Model: gpt-5-thinking
Create a Jobs-To-Be-Done interview script to uncover high-value problems for {{app_niche}}. Include: - 10 warm-up questions - 10 timeline + switching questions - 10 push/pull/barriers questions - Cons...
Author: Tsubasa Kato
Category: ideation | Model: gpt-5-thinking
Design a purely educational docking simulation plan with public PDB examples only. Deliver: objective, dataset list, feature extraction, baseline (shape complementarity), KPIs (rank@k), error analysis...
Author: Tsubasa Kato
Category: Simulation | Model: GPT-5 Thinking
Act as an AI rollout lead for a sub-50 person company. Deliver a 14-day plan: use cases ranked by ROI/risk, 2 quick-win agents (inbox triage, FAQ/RAG), minimal governance (human-in-the-loop), success ...
Author: Tsubasa Kato
Category: Strategy | Model: GPT-5 Thinking
Suggest hybrid quantum-classical approaches to optimize search ranking or graph traversal in large datasets.
Author: [email protected]
Category: Research & Innovation | Model: GPT-5, GPT-4o, Sonar