empire-eval / README.md
Lilambd's picture
Clarify dataset provenance and update cadence
1a0ca25 verified
|
Raw History Blame Contribute Delete
19.8 kB
metadata
license: cc-by-4.0
pretty_name: 'Lilambd Empire Eval: grounded questions over living public ledgers'
language:
  - en
  - ja
task_categories:
  - question-answering
  - table-question-answering
tags:
  - rag
  - evaluation
  - benchmark
  - agents
  - tabular
  - grounding
  - hallucination
  - abstention
  - contamination-free
size_categories:
  - 10K<n<100K
configs:
  - config_name: all
    default: true
    data_files:
      - split: dev
        path:
          - data/dev/japan-prefecture-facts.jsonl
          - data/dev/machine-economy-daily.jsonl
          - data/dev/security-daily.jsonl
          - data/dev/world-biodiversity-cells.jsonl
          - data/dev/world-earth.jsonl
          - data/dev/world-ground.jsonl
          - data/dev/world-hazard.jsonl
          - data/dev/world-marketplaces.jsonl
          - data/dev/world-prices.jsonl
          - data/dev/world-pulse-daily.jsonl
          - data/dev/world-science.jsonl
          - data/dev/world-seeds.jsonl
          - data/dev/world-signals.jsonl
          - data/dev/world-sound.jsonl
          - data/dev/worlds-021-hibe.jsonl
          - data/dev/worlds-022-dingongfi.jsonl
          - data/dev/worlds-023-nenshee.jsonl
          - data/dev/worlds-024-chortsin.jsonl
          - data/dev/worlds-025-ithu.jsonl
          - data/dev/worlds-026-deranhaa.jsonl
          - data/dev/worlds-027-takcho.jsonl
          - data/dev/worlds-028-kuyus.jsonl
          - data/dev/worlds-029-sei.jsonl
          - data/dev/worlds-030-dena.jsonl
          - data/dev/japan-dpc-hospitals.jsonl
          - data/dev/world-markets.jsonl
      - split: test
        path:
          - data/test/japan-prefecture-facts.jsonl
          - data/test/machine-economy-daily.jsonl
          - data/test/security-daily.jsonl
          - data/test/world-biodiversity-cells.jsonl
          - data/test/world-earth.jsonl
          - data/test/world-ground.jsonl
          - data/test/world-hazard.jsonl
          - data/test/world-marketplaces.jsonl
          - data/test/world-prices.jsonl
          - data/test/world-pulse-daily.jsonl
          - data/test/world-science.jsonl
          - data/test/world-seeds.jsonl
          - data/test/world-signals.jsonl
          - data/test/world-sound.jsonl
          - data/test/worlds-021-hibe.jsonl
          - data/test/worlds-022-dingongfi.jsonl
          - data/test/worlds-023-nenshee.jsonl
          - data/test/worlds-024-chortsin.jsonl
          - data/test/worlds-027-takcho.jsonl
          - data/test/worlds-028-kuyus.jsonl
          - data/test/worlds-030-dena.jsonl
          - data/test/japan-dpc-hospitals.jsonl
          - data/test/world-markets.jsonl
          - data/test/worlds-025-ithu.jsonl
          - data/test/worlds-026-deranhaa.jsonl
          - data/test/worlds-029-sei.jsonl
  - config_name: japan-dpc-hospitals
    data_files:
      - split: dev
        path: data/dev/japan-dpc-hospitals.jsonl
      - split: test
        path: data/test/japan-dpc-hospitals.jsonl
  - config_name: japan-prefecture-facts
    data_files:
      - split: dev
        path: data/dev/japan-prefecture-facts.jsonl
      - split: test
        path: data/test/japan-prefecture-facts.jsonl
  - config_name: machine-economy-daily
    data_files:
      - split: dev
        path: data/dev/machine-economy-daily.jsonl
      - split: test
        path: data/test/machine-economy-daily.jsonl
  - config_name: security-daily
    data_files:
      - split: dev
        path: data/dev/security-daily.jsonl
      - split: test
        path: data/test/security-daily.jsonl
  - config_name: world-biodiversity-cells
    data_files:
      - split: dev
        path: data/dev/world-biodiversity-cells.jsonl
      - split: test
        path: data/test/world-biodiversity-cells.jsonl
  - config_name: world-earth
    data_files:
      - split: dev
        path: data/dev/world-earth.jsonl
      - split: test
        path: data/test/world-earth.jsonl
  - config_name: world-ground
    data_files:
      - split: dev
        path: data/dev/world-ground.jsonl
      - split: test
        path: data/test/world-ground.jsonl
  - config_name: world-hazard
    data_files:
      - split: dev
        path: data/dev/world-hazard.jsonl
      - split: test
        path: data/test/world-hazard.jsonl
  - config_name: world-marketplaces
    data_files:
      - split: dev
        path: data/dev/world-marketplaces.jsonl
      - split: test
        path: data/test/world-marketplaces.jsonl
  - config_name: world-markets
    data_files:
      - split: dev
        path: data/dev/world-markets.jsonl
      - split: test
        path: data/test/world-markets.jsonl
  - config_name: world-prices
    data_files:
      - split: dev
        path: data/dev/world-prices.jsonl
      - split: test
        path: data/test/world-prices.jsonl
  - config_name: world-pulse-daily
    data_files:
      - split: dev
        path: data/dev/world-pulse-daily.jsonl
      - split: test
        path: data/test/world-pulse-daily.jsonl
  - config_name: world-science
    data_files:
      - split: dev
        path: data/dev/world-science.jsonl
      - split: test
        path: data/test/world-science.jsonl
  - config_name: world-seeds
    data_files:
      - split: dev
        path: data/dev/world-seeds.jsonl
      - split: test
        path: data/test/world-seeds.jsonl
  - config_name: world-signals
    data_files:
      - split: dev
        path: data/dev/world-signals.jsonl
      - split: test
        path: data/test/world-signals.jsonl
  - config_name: world-sound
    data_files:
      - split: dev
        path: data/dev/world-sound.jsonl
      - split: test
        path: data/test/world-sound.jsonl
  - config_name: worlds-021-hibe
    data_files:
      - split: dev
        path: data/dev/worlds-021-hibe.jsonl
      - split: test
        path: data/test/worlds-021-hibe.jsonl
  - config_name: worlds-022-dingongfi
    data_files:
      - split: dev
        path: data/dev/worlds-022-dingongfi.jsonl
      - split: test
        path: data/test/worlds-022-dingongfi.jsonl
  - config_name: worlds-023-nenshee
    data_files:
      - split: dev
        path: data/dev/worlds-023-nenshee.jsonl
      - split: test
        path: data/test/worlds-023-nenshee.jsonl
  - config_name: worlds-024-chortsin
    data_files:
      - split: dev
        path: data/dev/worlds-024-chortsin.jsonl
      - split: test
        path: data/test/worlds-024-chortsin.jsonl
  - config_name: worlds-025-ithu
    data_files:
      - split: dev
        path: data/dev/worlds-025-ithu.jsonl
      - split: test
        path: data/test/worlds-025-ithu.jsonl
  - config_name: worlds-026-deranhaa
    data_files:
      - split: dev
        path: data/dev/worlds-026-deranhaa.jsonl
      - split: test
        path: data/test/worlds-026-deranhaa.jsonl
  - config_name: worlds-027-takcho
    data_files:
      - split: dev
        path: data/dev/worlds-027-takcho.jsonl
      - split: test
        path: data/test/worlds-027-takcho.jsonl
  - config_name: worlds-028-kuyus
    data_files:
      - split: dev
        path: data/dev/worlds-028-kuyus.jsonl
      - split: test
        path: data/test/worlds-028-kuyus.jsonl
  - config_name: worlds-029-sei
    data_files:
      - split: dev
        path: data/dev/worlds-029-sei.jsonl
      - split: test
        path: data/test/worlds-029-sei.jsonl
  - config_name: worlds-030-dena
    data_files:
      - split: dev
        path: data/dev/worlds-030-dena.jsonl
      - split: test
        path: data/test/worlds-030-dena.jsonl
  - config_name: multilingual
    data_files:
      - split: en
        path: data/multilingual/en.jsonl
      - split: es
        path: data/multilingual/es.jsonl
      - split: ja
        path: data/multilingual/ja.jsonl
      - split: zh
        path: data/multilingual/zh.jsonl
  - config_name: world-ledgers
    data_files:
      - split: dev
        path: data/world-ledgers/dev.jsonl
      - split: test
        path: data/world-ledgers/test.jsonl

Lilambd Empire Eval

Question-answer evaluation sets grounded in pinned public ledger revisions. Sources include Lilambd benchmark generator and the source datasets linked by revision. Update cadence: Versioned evaluation sets; no fixed release interval stated. The dataset card lists CC BY 4.0 as its license. Questions and scripts are CC BY 4.0; answer licensing varies, including noncommercial-only rows.

6,456 questions whose answers you can check row by row. Each question is asked against one published Lilambd dataset. The answer was not written by a person or a model: it is what a small, machine-readable recipe returns when run on the dataset files, and every question names the rows that prove it (<file>#<row>).

Use it to measure whether a RAG pipeline, a table agent or a tool-using LLM actually reads the data, and whether it says "not in the data" when the data has no answer.

What makes it different

  • Every answer is re-derivable. scripts/verify.py reloads the dataset files, re-runs every recipe and re-reads every cited row. Result on 2026-11-03: 5,483 / 5,483 recipe answers reproduced exactly.
  • It tests abstention. 173 questions have no answer in the data (a future month, a region with no station, a fictional name). The scorer reports a hallucination rate next to accuracy.
  • It cannot be memorised. 10 of the sets are fictional planets (Lilambd Worlds) simulated from plate tectonics to languages; nothing about them exists on the web, so a correct answer must come from the files.
  • It renews itself. The real-world ledgers are rebuilt from public measurement records (daily for some), and the generator scripts are included, so a fresh question set can be cut from any later revision.
  • Held-out test split. 30% of the questions are published without answers. Their answers are kept by the maintainer so the split stays clean.

Sets

set source dataset dev test no-answer licence of answers
japan-dpc-hospitals Japan DPC Hospitals: all acute-care hospitals: beds, discharges, case mix 156 72 2 CC-BY-4.0
japan-prefecture-facts Japan by Prefecture: sourced facts for all 47 prefectures 136 70 0 CC-BY-4.0
machine-economy-daily Machine Economy Daily: x402 hosts, MCP registry, agent packages 81 40 2 CC-BY-4.0
security-daily Security Daily: CVE, KEV, EPSS, breaches, joined daily 140 62 6 CC-BY-4.0
world-biodiversity-cells World Biodiversity Cells: GBIF records and species on a 0.5° grid 181 72 0 CC-BY-NC-4.0
world-earth World Earth Surface Ledger: forest loss, land cover, built-up area, soil per region 220 110 18 CC-BY-4.0
world-ground World Ground Ledger: people near volcanoes, coasts and plate boundaries; soil 202 102 12 CC-BY-4.0
world-hazard World Hazard Ledger: earthquakes per region, station warming, disaster deaths 306 113 22 CC-BY-4.0
world-marketplaces World Marketplaces: digital marketplaces and 409k physical markets 250 101 5 CDLA-Permissive-2.0
world-markets World Markets: 138k physical markets, bazaars, night markets 161 55 0 CDLA-Permissive-2.0
world-prices World Prices Long Record: commodity prices since 1960, CPI, energy prices 217 101 20 CC-BY-4.0
world-pulse-daily World Pulse Daily: 15 public catalogues written down daily 235 94 2 CC-BY-SA-4.0
world-science World Science: OpenAlex paper counts by country, field, year 248 80 10 CC0-1.0
world-seeds World Seeds: Wikipedia "by country" tables keyed by ISO2 124 58 0 CC-BY-SA-4.0
world-signals World Signals: daily attention snapshot: Steam, GitHub, HF, packages, podcasts, search 164 58 2 CC-BY-4.0
world-sound World Sound: instruments, Cantometrics songs, tunings (dataset upload pending) 241 127 12 CC-BY-4.0
worlds-* (10 sets) Lilambd Worlds fictional planets: Hibe, Dingongfi, Nenshee, Chortsin, Ithu, Deranhaa, Takcho, Kuyus, Sei, Dena 1440 639 60 CC-BY-4.0
total 4,502 1,954 173

Task types: lookup 3,635, count 769, rank 735, aggregate 322, multi_hop 286, compare 193, reasoning 179, unanswerable 173, temporal 70, derived 47, nunique 37, documentation 10.

Skills tagged: table_lookup 2,537, contamination_free 2,079, free_text 973, count 769, filter 680, ranking 663, fresh_data 579, temporal 477, multilingual_entity 466, id_lookup 419, aggregation 293, code_resolution 246, comparison 193, security 191.

One row

{
 "id": "world-hazard-0001",
 "dataset": "Lilambd/world-hazard",
 "revision": "hf-main@2026-11-03 (data as of 2026-11-01)",
 "task": "lookup",
 "question": "In the World Hazard Ledger, how many M6+ earthquakes from 1900 to 2026 are assigned to Eastern Samar (Philippines)?",
 "answer": "26",
 "answer_type": "number",
 "tolerance": {
  "abs": 0.0,
  "rel": null
 },
 "evidence": [
  {
   "row_id": "data/admin1_quake.csv#92",
   "file": "data/admin1_quake.csv",
   "row": 92,
   "key": "{\"a1\": \"PHL-2582\", \"name\": \"Eastern Samar\", \"country\": \"Philippines\"}"
  }
 ],
 "recipe": "{\"op\": \"lookup\", \"file\": \"data/admin1_quake.csv\", \"where\": [[\"name\", \"==\", \"Eastern Samar\"], [\"country\", \"==\", \"Philippines\"]], \"col\": \"m6\"}"
}

Fields: answer (text form), answer_json (typed value: number, string, list, boolean or null for no answer), answer_type, tolerance (abs or rel for numbers), evidence (row ids with their key columns), evidence_scope (the filter behind counts, sums and rankings), recipe (JSON: the operation that produces the answer), skills, difficulty 1-3, revision (dataset revision the answer belongs to).

Multilingual subset

data/multilingual/ holds the same 1,000 dev questions in en, es, ja, zh. Only the question text is translated; names, codes, numbers, column names and quoted strings are kept exactly, and the answer, evidence and recipe are the English item's. Each row keeps source_id and question_en, so a gap between languages is a reading gap, not a data gap. All recipe answers in every language file were re-derived by verify.py. Use the multilingual config.

World Ledgers (SQL recipes)

The world-ledgers config adds 9,999 questions (dev 7,076, test 2,923, same hash split) over the Lilambd world ledgers of 2026-11-03: borders, calendars, cities, climate, currency, demand, food, languages, objects, population, power. 4,179 of them join two or three ledgers (city_id, ISO codes, nearest city). Here the recipe is a DuckDB query (recipe_sql) and the evidence is the list of files it reads (evidence_files); question_ja gives a Japanese wording. scripts/verify_sql.py --views scripts/world_ledgers_views.json re-runs every query; on 2026-11-03 it reproduced 9,999 / 9,999 answers (one item whose query no longer matched the ledger files was removed). Not every ledger file is on the Hub yet, so outside users can score against these answers but cannot all re-run the queries today.

Licence: 1,652 of these rows (commercial_ok: false) rest on CShapes (non-commercial) or FAOSTAT production data; use them for non-commercial evaluation only. The rest follow CC BY 4.0 with source attribution (licence_note).

Score a system

pip install pandas pyarrow pycountry
python scripts/make_prompts.py --qa data/dev --out prompts.jsonl --with-context-hint
# run your system, write {"id": ..., "answer": ...} per line to predictions.jsonl, or use the bundled runner:
ANTHROPIC_API_KEY=... python scripts/run_eval.py --qa data/dev --provider anthropic --out results/claude/predictions.jsonl --concurrency 8
OPENAI_API_KEY=... python scripts/run_eval.py --qa data/dev --provider openai --base-url http://localhost:8000/v1 --model MODEL --out results/MODEL/predictions.jsonl
python scripts/score.py --qa data/dev --pred predictions.jsonl --out score.json

The scorer accepts numbers with commas or units, lists as JSON or comma-separated, and treats null, empty, "not in the dataset", "unknown" and similar as abstaining. It reports accuracy, answerable accuracy, abstention accuracy, hallucination rate and false-abstention rate, broken down by set, task, difficulty and skill. Use --lenient for free-text answers (the fictional-planet questions written by the world generator).

Baselines on dev (from results/baseline_*/score.json): oracle (reads the answers) accuracy 1.0000, hallucination rate 0.00; always abstain accuracy 0.0362, hallucination rate 0.00; random guess accuracy 0.0003, hallucination rate 1.00. leaderboard/build_leaderboard.py turns any number of score.json files into leaderboard.json and a static leaderboard.html; leaderboard/EVAL_CARD.md has a model-index snippet for reporting scores on a model card.

Test split: send predictions.jsonl for the test ids to submissions@lilambd.com and you get the same report back.

Rebuild or extend

scripts/fetch_hf.py downloads the source files pinned to their current commit, scripts/build.py generates questions (fixed seed), scripts/verify.py re-derives them, scripts/package.py makes the dev/test split. A private set over your own tables can be built the same way.

Limits

  • Answers belong to the stated revision. The daily ledgers move on; re-verify against the same revision, or rebuild.
  • Many questions name a table or a column (for example features_country.csv, m6_in). They test grounded reading, not trivia.
  • Counting and ranking questions over large tables (hundreds of thousands of rows) are meant for systems that can run code or SQL; pure retrieval will not see every row.
  • About 970 fictional-planet questions come from the world generator's own question list: free-text answers, checked by the generator's consistency tests, not by recipes (recipe is empty, verified_by says so).
  • In the worlds-* sets, source_qid (for example 021-000) is the world generator's internal question id, not a Wikidata QID.
  • Source keys that repeat with different values (31 country CPI series in world-prices monthly.csv, 26 region rows in world-earth features_region.csv) are left out of question generation, so no question has two possible answers.
  • Entity names are as written in the source; country names in questions come from ISO 3166 where the data has only a code.

Licence

Questions, recipes and scripts: CC BY 4.0, Lilambd. Answers are facts taken from the source datasets and keep their licence (column above): most are CC BY 4.0, World Science is CC0, the market sets are CDLA-Permissive-2.0. Cite the source dataset alongside this one.

@misc{lilambd_empire_eval_2026,
  title  = {Lilambd Empire Eval: grounded questions over living public ledgers},
  author = {Lilambd},
  year   = {2026},
  url    = {https://huggingface.co/datasets/Lilambd/empire-eval}
}

Need an evaluation set over your own data, a private held-out split, or a set in another language? submissions@lilambd.com


Missing something? Tell us in one line what you'd gladly pay for, if someone made it — one-line form: https://docs.google.com/forms/d/1tzMQtHsEbMg7JyMk7YGqslHijfNTciByi8wrwjQpUPY/viewform (or email submissions@lilambd.com, subject "Data wish"). If enough people ask, we build it, and you get it free.

Used this? Tell us what you used it for, what was missing, and what you used before: https://docs.google.com/forms/d/1tzMQtHsEbMg7JyMk7YGqslHijfNTciByi8wrwjQpUPY/viewform (one minute, no email asked).

Need a specific file or dataset made? Ask here (no promise, quick form): https://docs.google.com/forms/d/e/1FAIpQLSf6V7YALfwE42SaF1re_cLzAoP96TpRg61jE3Nf7UTWEtWRxA/viewform