{"success":true,"data":{"author":"Google DeepMind (Kajic, Wiles, Albuquerque, Bauer, Wang, Pont-Tuset, Nematzadeh)","bucketAccessType":"public","bucketListable":true,"bucketName":"geckonum_t2i_benchmark","bucketPrefix":"","category":"Computer Vision","cloudProvider":"gcp","createdAt":"2026-11-22T01:11:45.436Z","description":"GeckoNum evaluates whether text-to-image generative models draw the right number of objects. Its text prompts probe different aspects of numerical reasoning, from a single counted noun to prompts with two or three counted entities that must add up, with numbers written as words or as digits, grouped into 12 prompt types. The bucket holds the generated images, one folder per model (dalle_3, imagen_a, imagen_b, imagen_c, imagen_d, muse_a, muse_b), with PNG files named geckonum_<prompt_id>_<seed>.png for five seeds per prompt, plus geckonum.zip (18.8 GB) with all images and browsable index.html pages. The text prompts (prompts.csv) and the human annotations for three rating tasks (CSV per task and model) are in the GeckoNum GitHub repository and join to the images by prompt id. Used to benchmark object counting in image generation, compare automatic evaluation metrics with human judgments, and train or test text-image alignment scorers. License: CC BY 4.0 for the materials (software Apache 2.0), stated in the repository. Cite Kajic et al., Evaluating Numerical Reasoning in Text-to-Image Models (NeurIPS 2024).","displayTitle":"GeckoNum: Numerical Reasoning Benchmark for Text-to-Image Models, Generated Images from Imagen, Muse and DALL-E 3 with Counting Prompts (Google DeepMind)","fuseMountBrowsable":true,"id":"047cb975-f8f3-5880-b538-8450520bf8a6","license":"CC-BY-4.0","licenseEvidenceUrl":"https://github.com/google-deepmind/geckonum_benchmark_t2i","objectCountEstimate":8,"probeStatus":"ok","region":"us","schemaCard":{"accessHints":[{"language":"python-gcsfs","snippet":"import gcsfs\nfs = gcsfs.GCSFileSystem(token='anon')\nfs.ls('geckonum_t2i_benchmark/', detail=False)[:10]"},{"language":"python-storage","snippet":"from google.cloud import storage\nclient = storage.Client.create_anonymous_client()\nlist(client.bucket('geckonum_t2i_benchmark').list_blobs(prefix='', max_results=10))"}],"caveats":["Files are binary — use a domain-specific parser rather than text I/O."],"format":"binary","generatedAt":"2026-11-22T01:11:51.572Z","layout":{"example":"css/","pattern":"{id}/{object}"},"mountPath":"/mnt/gcs-archive/data","sampleHeaders":[{"bytes":"html {\n  height: 100%;\n}\nhtml,\nbody {\n  padding: 0;\n  margin: 0;\n}\nbody {\n  background: #fff;\n  color: #444;\n  font-family: 'Open Sans', sans-serif;\n  font-size: 16px;\n  font-weight: lighter;\n  margin-bottom: 1em;\n}\na {\n  text-decoration: none;\n}\n\n#container {\n  padding: 2em;\n}\nheader {\n  background: #fff;\n  color: #444;\n  font-size: 1em;\n  margin-bottom: 1.1em;\n}\nh1 {\n  display: inline-block;\n  font-weight: light;\n  vertical-align: top;\n}\nh1 a {\n  color: #444;\n}\n\n#media li {\n  float: left;\n  margin-right: ","encoding":"utf8","key":"css/","truncatedAt":512},{"bytes":"html {\n  height: 100%;\n}\nhtml,\nbody {\n  padding: 0;\n  margin: 0;\n}\nbody {\n  background: #fff;\n  color: #444;\n  font-family: 'Open Sans', sans-serif;\n  font-size: 16px;\n  font-weight: lighter;\n  margin-bottom: 1em;\n}\na {\n  text-decoration: none;\n}\n\n#container {\n  padding: 2em;\n}\nheader {\n  background: #fff;\n  color: #444;\n  font-size: 1em;\n  margin-bottom: 1.1em;\n}\nh1 {\n  display: inline-block;\n  font-weight: light;\n  vertical-align: top;\n}\nh1 a {\n  color: #444;\n}\n\n#media li {\n  float: left;\n  margin-right: ","encoding":"utf8","key":"css/style.css","truncatedAt":512}],"sampleObjects":["css/","css/style.css"],"sourceObjectKey":"css/","starterCell":"from google.cloud import storage\nclient = storage.Client.create_anonymous_client()\nbucket = client.bucket('geckonum_t2i_benchmark')\nblob = bucket.blob('css/')\nwith blob.open(\"rb\") as fh:\n    print(fh.read(64))","starterSamplePath":"css/style.css","totalObjects":"8 top-level prefixes","verifiedReadAttempt":{"at":"2026-11-22T06:16:14.472Z","attempts":[{"family":"text-head","ok":true,"path":"css/style.css","seconds":0},{"family":"bytes-head","ok":true,"path":"css/style.css","seconds":0},{"error":"ModuleNotFoundError: No module named 'xarray'","family":"netcdf-xarray","ok":false,"path":"css/style.css","seconds":0},{"family":"directory","ok":true,"path":"css/","seconds":0.03}],"facts":[],"image":"531775725781.dkr.ecr.us-east-1.amazonaws.com/scigantic-notebook-minimal:latest","outcome":"no-reader","starterCell":{"error":"ModuleNotFoundError: No module named 'google'","ok":false}}},"summary":"GeckoNum evaluates whether text-to-image generative models draw the right number of objects.","title":"GeckoNum: Numerical Reasoning Benchmark for Text-to-Image Models, Generated Images from Imagen, Muse and DALL-E 3 with Counting Prompts (Google DeepMind)","updatedAt":"2026-11-22T04:06:41.968Z","visibility":"public","websiteUrl":"https://github.com/google-deepmind/geckonum_benchmark_t2i"}}