RISE Humanities Data Benchmark, 0.5.5-pre1

Search Test Runs

 

A test run is a single execution of a benchmark test using a defined model configuration.
Each run represents how a particular large language model (LLM) — such as GPT-4, Claude-3, or Gemini — performed on a given task at a specific time, with specific settings.

A test run includes:

  • Prompt and role definition – what the model was asked to do and from what perspective (e.g. “as a historian”).
  • Model configuration – provider, model version, temperature, and other generation parameters.
  • Results – the model’s actual response and its evaluation (scores such as F1 or accuracy).
  • Usage and cost data – token counts and calculated API costs.
  • Metadata – information like the test date, benchmark name, and person who executed it.

Together, test runs make it possible to compare models, providers, and configurations across benchmarks in a transparent and reproducible way.

Search Results

Your search for Benchmark 'general_meeting_minutes__true' with Search Hidden 'False' returned 92 results, showing page 1 of 10.
Result 1 of 92

Test T1643 at 2026-08-18

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Provideropenrouter
Modelz-ai/glm-5v-turbo
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score75.10 %
Test time70.11 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_23_Assemblée_1951_Generalversammlung1_page_3","page_number":3,"entries":[{"number":"21","name":"M.de la FONTAINE Lionel","address":"Boîte postale 41 à MARRAKECH-MEDINA","actions_o":"","actions_p":"2","no_de_voix":"","signature_present":false,"signature":""},{"number":"22","name":"M.LAUMAIN Robert","address":"24 Avenue Gambetta MONTROUGE","actions_o":"","actions_p":"1","no_de_voix":"","signature_present":false,"signature":"Aut. Pau"},{"number":"23","name":"Suc.IHOEST Paul (Mme)","address":"1,rue J.B.Colyns IXELLES-BRUXELLES.","actions_o":"150","actions_p":"","no_de_voix":"","signature_present":false,"signature":""},{"number":"24","name":"M.MASSON Daniel","address":"3,Av.Gaugé VIROFLAY (S-et-O)","actions_o":"","actions_p":"10","no_de_voix":"","signature_present":false,"signature":"Aut. Pau"},{"number":"25","name":"M. PANI Antonio","address":"SESSA (Tessin) Suisse","actions_o":"1","actions_p":"10","no_de_voix":"","signature_present":false,"signature":"Aut. Pau"},{"number":"26","name":"M. POUQUET Maurice","address":"Essendières,par St Médard d'Excisuil (Dordogne)","actions_o":"","actions_p":"1","no_de_voix":"","signature_present":false,"signature":""},{"number":"27","name":"Mme RAOULT Suzanne","address":"195 rue de Grenelle PARIS","actions_o":"","actions_p":"10","no_de_voix":"","signature_present":false,"signature":""},{"number":"28","name":"Suc.RICHOMME de POSSON (Mme)","address":"37bis Promenade des Anglais NICE (A-M)","actions_o":"","actions_p":"32","no_de_voix":"","signature_present":false,"signature":""},{"number":"29","name":"Mme SAVAEY M.L.","address":"5,Avenue Weber GENEVE (Suisse)","actions_o":"","actions_p":"10","no_de_voix":"","signature_present":false,"signature":""},{"number":"30","name":"M. SCHERESCHWESKY Philippe","address":"34,Bd Marbeau PARIS (16°)","actions_o":"","actions_p":"1","no_de_voix":"","signature_present":false,"signature":""}],"total_actions":{"total_o":"2463","total_p":"856","total_voix":""}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.75 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 41.1K IT + 8.2K OT = 49.3K TTCost: 0.049$0.033$0.082$
Result 2 of 92

Test T1598 at 2026-08-18

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providergenai
Modelgemini-3.7-flash
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score93.69 %
Test time6.28 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_19_Assemblée Mine d. Costano_1_Assemblee 28.12.1967_1_page_4","page_number":4,"entries":[{"number":"I","name":"Monsieur BREHAM Alain","address":"45, bd du Commandant Charcot
NEUILLY S/SEINE","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":false,"signature":""},{"number":"2","name":"Monsieur BREHAM Marc","address":"La Tuillerie
SAINT NOM LA BRETECHE","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":false,"signature":""},{"number":"3","name":"Monsieur MASSON Daniel","address":"Villa La Cava Avenue de La Cava
06 / CANNES","actions_o":"IO","actions_p":"","no_de_voix":"IO","signature_present":false,"signature":""},{"number":"4","name":"Compagnie Centrale SICLI","address":"II bis, rue Balzac
PARIS 8e","actions_o":"I68","actions_p":"822","no_de_voix":"990","signature_present":false,"signature":""},{"number":"5","name":"Société des Mines de l'Aurieras","address":"II bis, rue Balzac
PARIS 8e","actions_o":"45","actions_p":"355","no_de_voix":"400","signature_present":false,"signature":""},{"number":"6","name":"Société des Recherches Minières Françaises","address":"II bis, rue Balzac
PARIS 8e","actions_o":"IO","actions_p":"","no_de_voix":"IO","signature_present":false,"signature":""},{"number":"7","name":"Monsieur PANI Antonio","address":"SESSA - TESSIN
SUISSE","actions_o":"IO","actions_p":"","no_de_voix":"IO","signature_present":false,"signature":""}],"total_actions":{"total_o":"","total_p":"","total_voix":""}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.94 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 10.8K IT + 9.1K OT = 20.0K TTCost: 0.008$0.034$0.042$
Result 3 of 92

Test T1628 at 2026-08-18

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Provideropenrouter
Modelmeta/muse-spark-1.2
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score91.72 %
Test time16.29 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_23_Assemblée_1951_Generalversammlung5_page_20","page_number":20,"entries":[{"number":"31","name":"M. SIRCOULON Jacques","address":"Mesli\u0000e8res (Doubs)","actions_o":"11","actions_p":"","no_de_voix":"","signature_present":false,"signature":""},{"number":"32","name":"St\u0000e9 Anonyme des Mines d'OR du CHATELET","address":"11bis rue Balzac
PARIS (8\u0000b0)","actions_o":"15","actions_p":"365","no_de_voix":"","signature_present":true,"signature":"S. Feur [mandataire]"},{"number":"33","name":"M. STRAP Gaston","address":"112 Bd Exelmans PARIS (16\u0000b0)","actions_o":"64","actions_p":"50","no_de_voix":"","signature_present":true,"signature":"S. Feur [mandataire]"},{"number":"34","name":"M. STRAP Pierre","address":"73, rue des Vignes PARIS (16\u0000b0)","actions_o":"1","actions_p":"","no_de_voix":"","signature_present":true,"signature":"Aut. Paur"},{"number":"35","name":"Mme WARNANT Pierre","address":"148 Av. Franklin Roosevelt
BRUXELLES (Belgique)","actions_o":"","actions_p":"200","no_de_voix":"","signature_present":false,"signature":""},{"number":"36","name":"de ZANNA Arcangelo","address":"","actions_o":"","actions_p":"1","no_de_voix":"","signature_present":false,"signature":""}],"total_actions":{"total_o":"1010","total_p":"3200","total_voix":""}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.92 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 37.5K IT + 27.6K OT = 65.1K TTCost: 0.047$0.117$0.164$
Result 4 of 92

Test T1613 at 2026-08-18

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerx-ai
Modelgrok-4.6
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score92.21 %
Test time34.84 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_18_Assemblées_3_Bilanz1968_2_page_9","page_number":9,"entries":[{"number":"1","name":"Monsieur BREHAM Alain","address":"45 bd Charcot - NEUILLY-S/SEINE (92)
FRANCE","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":false,"signature":""},{"number":"2","name":"Monsieur BREHAM Marc","address":"la Tuilerie - ST NOM LA BRETECHE
FRANCE","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":false,"signature":""},{"number":"3","name":"Monsieur MASSON Daniel","address":"Villa la Cava, av. de la Cava 06 -
FRANCE CANNES","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":false,"signature":""},{"number":"4","name":"COMPAGNIE CENTRALE SICLI","address":"2/4 rue Blaise Pascal
93 - LE BLANC-MESNIL FRANCE","actions_o":"168","actions_p":"822","no_de_voix":"990","signature_present":false,"signature":""},{"number":"5","name":"SOCIETE DES MINES DE L'AURIERAS","address":"11 bis rue Balzac PARIS 8e
FRANCE","actions_o":"45","actions_p":"355","no_de_voix":"400","signature_present":false,"signature":""},{"number":"6","name":"SOCIETE DE RECHERCHES MINIERES FRAN-
CAISES","address":"11 bis rue Balzac PARIS 8e
FRANCE","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":false,"signature":""},{"number":"7","name":"Monsieur PANI Antonio","address":"SESSA (Tessin)
SUISSE","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":false,"signature":""}],"total_actions":{"total_o":"","total_p":"","total_voix":""}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.92 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 28.5K IT + 7.5K OT = 36.1K TTCost: 0.057$0.045$0.102$
Result 5 of 92

Test T1583 at 2026-08-16

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerhuggingface
Modelmeta-models/Muse-Glimmer-30B
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score85.87 %
Test time26.85 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_18_Assemblées_3_Bilanz1968_2_page_9","page_number":9,"entries":[{"number":"1","name":"Monsieur BREHAM Alain","address":"45 bd Charcot - NEUILLY-S/SEINE (92)
FRANCE","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":false,"signature":""},{"number":"2","name":"Monsieur BREHAM Marc","address":"la Tuilerie - ST NOM LA BRETECHE
FRANCE","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":false,"signature":""},{"number":"3","name":"Monsieur MASSON Daniel","address":"Villa la Cava, av. de la Cava 06 -
CANNES
FRANCE","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":false,"signature":""},{"number":"4","name":"COMPAGNIE CENTRALE SICLI","address":"2/4 rue Blaise Pascal
93 - LE BLANC-MESNIL
FRANCE","actions_o":"168","actions_p":"822","no_de_voix":"990","signature_present":false,"signature":""},{"number":"5","name":"SOCIETE DES MINES DE L'AURIERAS","address":"11 bis rue Balzac PARIS 8e
FRANCE","actions_o":"45","actions_p":"355","no_de_voix":"400","signature_present":false,"signature":""},{"number":"6","name":"SOCIETE DE RECHERCHES MINIERES FRANCAISES","address":"11 bis rue Balzac PARIS 8e
FRANCE","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":false,"signature":""},{"number":"7","name":"Monsieur PANI Antonio","address":"SESSA (Tessin)
SUISSE","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":false,"signature":""}],"total_actions":{"total_o":"","total_p":"","total_voix":""}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.86 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 36.8K IT + 25.6K OT = 62.4K TTCost: 0.013$0.038$0.051$
Result 6 of 92

Test T1553 at 2026-08-16

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerhuggingface
Modelthinkingmachines/Inkling-Small
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score75.27 %
Test time42.07 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_19_Assemblée Mine d. Costano_2.3_27.12.1961.pdf","page_number":3,"entries":[{"number":"1","name":"M. BREHAM Alain","address":"45 Boulevard Charcot - Neuilly-sur-Seine (Seine)","actions_o":"520","actions_p":"170","no_de_voix":"170","signature_present":true,"signature":"P. Pon"},{"number":"2","name":"M. BREHAM Marc","address":"16, rue Saint Pierre - Neuilly sur Seine (Seine)","actions_o":"520","actions_p":"170","no_de_voix":"170","signature_present":true,"signature":"P. Pon"},{"number":"3","name":"M. MASSON Daniel","address":"9 rue de l'orangerie Compiègne (Oise)","actions_o":"-","actions_p":"-","no_de_voix":"10","signature_present":true,"signature":"P. Pon"},{"number":"4","name":"Cie Centrale des mines et de métallurgie Sicilienne","address":"11, rue Balzac - Paris (8°)","actions_o":"222","actions_p":"162","no_de_voix":"162","signature_present":true,"signature":"P. Pon"},{"number":"5","name":"M. STRAP Pierre","address":"73, rue des Vignes - Paris XVIe","actions_o":"80","actions_p":"68","no_de_voix":"68","signature_present":true,"signature":"P. Pon"},{"number":"6","name":"Société des mines de l'AURIERAS","address":"11, rue Balzac - Paris (8°)","actions_o":"355","actions_p":"45","no_de_voix":"45","signature_present":true,"signature":"P. Pon"},{"number":"7","name":"Société des Recherches Minières Françaises","address":"11, rue Balzac - Paris (8°)","actions_o":"-","actions_p":"-","no_de_voix":"10","signature_present":true,"signature":"P. Pon"},{"number":"8","name":"M. PANL Antonio","address":"Sessa (Tessin)","actions_o":"-","actions_p":"-","no_de_voix":"10","signature_present":true,"signature":"P. Pon"},{"number":"9","name":"dott. Travaso","address":"Paris","actions_o":"1","actions_p":"1","no_de_voix":"1","signature_present":false,"signature":""},{"number":"10","name":"Paul Marx","address":"","actions_o":"1","actions_p":"1","no_de_voix":"1","signature_present":false,"signature":""},{"number":"11","name":"Hi Beatrice","address":"","actions_o":"1","actions_p":"1","no_de_voix":"1","signature_present":false,"signature":""}],"total_actions":{"total_o":"2999","total_p":"1599","total_voix":"1599"}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.75 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 19.5K IT + 42.7K OT = 62.2K TTCost: 0.009$0.051$0.060$
Result 7 of 92

Test T1493 at 2026-08-15

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerhuggingface
Modelswiss-ai/apertus-v1.5-70b
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score62.01 %
Test time199.68 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_23_Assemblée_1951_Generalversammlung1_page_2","page_number":2,"entries":[{"number":"1","name":"M. SIEMENSCHI, Edgard","address":"15, rue de l'Abbaye","actions_o":"3","actions_p":"1","no_de_voix":"4","signature_present":true,"signature":"Signature"},{"number":"2","name":"M. JEANNOT, Paul","address":"45, rue de la Station (Schaerbeek)","actions_o":"2","actions_p":"1","no_de_voix":"3","signature_present":true,"signature":"Signature"},{"number":"3","name":"M. RENIER, Paul","address":"4, rue Achille Devos (Schaerbeek)","actions_o":"3","actions_p":"1","no_de_voix":"4","signature_present":true,"signature":"Signature"},{"number":"4","name":"M. BOVENKENS, Georges","address":"5, rue Basse (446)","actions_o":"4","actions_p":"1","no_de_voix":"5","signature_present":true,"signature":"Signature"},{"number":"5","name":"M. STEVENS, Georges","address":"6, rue Basse (448)","actions_o":"5","actions_p":"1","no_de_voix":"6","signature_present":true,"signature":"Signature"},{"number":"6","name":"M. STEVENS, André","address":"10, av. de la Jonction (446)","actions_o":"6","actions_p":"1","no_de_voix":"7","signature_present":true,"signature":"Signature"},{"number":"7","name":"M. SERVAIS, André","address":"59, av. Général de Gaulle (Schaerbeek)","actions_o":"7","actions_p":"1","no_de_voix":"8","signature_present":true,"signature":"Signature"},{"number":"8","name":"M. DE GROOTE, André","address":"7, quai du Molenbeek (Schaerbeek)","actions_o":"8","actions_p":"1","no_de_voix":"9","signature_present":true,"signature":"Signature"},{"number":"9","name":"M. BROUX, André","address":"7, quai du Molenbeek (Schaerbeek)","actions_o":"9","actions_p":"1","no_de_voix":"10","signature_present":true,"signature":"Signature"}],"total_actions":{"total_o":"31","total_p":"9","total_voix":"45"}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.62 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 73.8K IT + 7.6K OT = 81.3K TTCost: 0.060$0.022$0.083$
Result 8 of 92

Test T1538 at 2026-08-15

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerhuggingface
ModelMiniMaxAI/MiniMax-M3
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score74.92 %
Test time13.54 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_18_Assemblées_4_Bilanz1965_1_page_7","page_number":7,"entries":[{"number":"1","name":"M. BREHAM Alain","address":"15 Bd Charcot Neuilly s/Seine France","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":true,"signature":"J. Périer"},{"number":"2","name":"M. BREHAM Marc La Tuilière","address":"Saint-Nom la Bretèche (78) France","actions_o":"179","actions_p":"520","no_de_voix":"699","signature_present":true,"signature":"J. Périer"},{"number":"3","name":"M. MASSON Daniel","address":"9 Rue de l'Hourgerie Courbevoie (60) France","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":true,"signature":"M. Périer"},{"number":"4","name":"Compagnie Centrale SICLI","address":"11/13 Rue Balzac Paris (8e)","actions_o":"162","actions_p":"822","no_de_voix":"990","signature_present":true,"signature":"J. Périer"},{"number":"5","name":"Société MS Paris et Aumône","address":"11/13 Rue Balzac Paris (8e)","actions_o":"45","actions_p":"355","no_de_voix":"400","signature_present":true,"signature":"J. Périer"},{"number":"6","name":"Société de Recherche Minière Francemines","address":"11/13 Rue Balzac Paris (8e)","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":true,"signature":"J. Périer"},{"number":"7","name":"M. le Professeur PBNI Antoine","address":"Sessy (T)","actions_o":"10","actions_p":"","no_de_voix":"10","signature_present":true,"signature":"J. Périer"}],"total_actions":{"total_o":"601","total_p":"2217","total_voix":""}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.75 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 7.6K IT + 8.1K OT = 15.7K TTCost: 0.002$0.009$0.011$
Result 9 of 92

Test T1523 at 2026-08-15

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerhuggingface
ModelQwen/Qwen3-VL-235B-A22B-Instruct
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score65.46 %
Test time160.09 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

{"document":"Se_18_Assemblées_4_Bilanz1965_1_page_7","page_number":7,"entries":[{"number":"~/1","name":"M. BREHAM Alain","address":"45 Bd Chancot Neuilly s/Seine - France","actions_o":"~/179","actions_p":"~/520","no_de_voix":"~/699","signature_present":true,"signature":"A. Pau"},{"number":"~/2","name":"M. BREHAM Marc - La Tuilerie","address":"Saint-Nom la Bretèche (78) France","actions_o":"~/179","actions_p":"~/520","no_de_voix":"~/699","signature_present":true,"signature":"Te. Pau"},{"number":"~/3","name":"M. MASSON Daniel","address":"9 Rue de l'Orangerie Couvreligne (60)
France.","actions_o":"~/10","actions_p":"~/0","no_de_voix":"~/10","signature_present":true,"signature":"Mara Pau"},{"number":"~/4","name":"Compagnie Centrale Sicli","address":"116a Rue Balzac - Paris (8e)","actions_o":"~/168","actions_p":"~/822","no_de_voix":"~/990","signature_present":true,"signature":"Ta Pau"},{"number":"~/5","name":"Société des Cimes del-Aurieras","address":"116a Rue Balzac - Paris (8e)","actions_o":"~/45","actions_p":"~/355","no_de_voix":"~/400","signature_present":true,"signature":"B. He"},{"number":"~/6","name":"Société de Recherches Minières Françaises","address":"116a Rue Balzac - Paris (8e)","actions_o":"~/10","actions_p":"~/0","no_de_voix":"~/10","signature_present":true,"signature":"A. Pau"},{"number":"~/7","name":"M. le Professeur BANI Antonio","address":"Jessa (Ti)","actions_o":"~/10","actions_p":"~/0","no_de_voix":"~/10","signature_present":true,"signature":"A. Pau"}],"total_actions":{"total_o":"~/601","total_p":"~/2217","total_voix":"~/2818"}}

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.65 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 28.9K IT + 9.5K OT = 38.4K TTCost: 0.006$0.008$0.014$
Result 10 of 92

Test T1568 at 2026-08-15

minutes information-extraction typed, handwritten 20 it, fr, de table person, location

Configuration
Providerhuggingface
Modelthinkingmachines/Inkling
  
Temperature0.0
DataclassMinutesPage
  
Normalized Score9.29 %
Test time51.10 seconds
Prompt

Please extract the metadata according to the given output format.
Name and Address are in the same table field. Your task is to extract them into separate fields.
Lose any dashes between name and address but preserve linebreaks.
Be aware: Addresses may contain multiple dashes; preserve them, only remove "visual splitting characters" between name and address.
Filename: {filename}
Page: {page_number}

Results

We need parse image table. Need extract entries 1-7 and totals at bottom. Need separate name and address, removing dashes that split them but preserve line breaks and internal dashes in address.

Let's read table carefully.

Columns: N° D'ORDRE, NOM, PRÉNOMS ET DOMICILE DES ACTIONNAIRES, ACTIONS (Ordinaire, de Priorité, DE VOIX), NOMBRE DE VOIX? Actually columns: Ordinaire, de Priorité? Wait "ACTIONS" with subcols "Ordi- naires" and "de Prio- rité" and "DE VOIX". Then "NOMBRE" and "SIGNATURE DES ACTIONNAIRES OU DES MANDATAIRES".

Data rows:
1: Mr Alain BREHAM (Cabinet D. FEAU - 132 Boulevard Haussmann - 75 / PARIS 8? Actually address line split: "132 Boulevard Haussmann - 75 / PARIS 8" maybe "132 Boulevard Haussmann - 75 / PARIS 8"? Need separate name and address. Name: "Mr Alain BREHAM (Cabinet D. FEAU"?? Wait there is dash between name and address? The text in cell: "Mr Alain BREHAM (Cabinet D. FEAU - 132 Boulevard Haussmann - 75 / PARIS 8" Hmm need understand splitting characters. The instruction: "Lose any dashes between name and address but preserve linebreaks. Be aware: Addresses may contain multiple dashes; preserve them, only remove 'visual splitting characters' between name and address."

So in row 1, the cell contains:
Mr Alain BREHAM (Cabinet D. FEAU -
132 Boulevard Haussmann - 75 / PARIS 8

Thus the dash after "FEAU" is the visual splitter between name and address? Actually name includes "(Cabinet D. FEAU" maybe that's part of name/organization? Let's parse: "Mr Alain BREHAM (Cabinet D. FEAU" then dash then address "132 Boulevard Haussmann - 75 / PARIS 8". So name should be "Mr Alain BREHAM (Cabinet D. FEAU" and address "132 Boulevard Haussmann - 75 / PARIS 8". Need remove the dash that splits them, but preserve line break. So name: "Mr Alain BREHAM (Cabinet D. FEAU)"? Wait there is opening parenthesis but no closing? In image maybe "(Cabinet D. FEAU -" then address. Actually might be "Mr Alain BREHAM (Cabinet D. FEAU)" as name, then address on next line. But we just need separate fields with line breaks preserved? The address field should contain the address lines with line breaks? The schema uses string, so we can include newline characters? Probably yes preserve linebreaks. Let's represent with 
? But JSON can include actual newline escaped. We'll use 
 for clarity? Actually in JSON string we can embed 
.

Let's list all rows precisely.

Row 1:
Number: 1
Actions Ordinaire: 179
Actions Priorité: 520
Voix: 699
Name/Address cell:
Line1: Mr Alain BREHAM (Cabinet D. FEAU -
Line2: 132 Boulevard Haussmann - 75 / PARIS 8
So name = "Mr Alain BREHAM (Cabinet D. FEAU"
Address = "132 Boulevard Haussmann - 75 / PARIS 8"
But need lose dashes between name and address. The dash after FEAU is removed. Also maybe there is a line break. So address starts on new line.

Row 2:
Number: 2
Name/Address:
Mr Marc BREHAM - La Tuilerie -
78/St-Nom la Bretèche
Wait need split name and address. The dash after BREHAM is splitter? Then "La Tuilerie" might be part of address? Actually "Mr Marc BREHAM - La Tuilerie -" then address "78/St-Nom la Bretèche". Let's parse: Name "Mr Marc BREHAM", then splitter dash, then address "La Tuilerie -
78/St-Nom la Bretèche"? Or "La Tuilerie" is part of address? The instruction says addresses may contain multiple dashes; preserve them, only remove visual splitting characters between name and address. Here there are dashes after BREHAM and after Tuilerie. Which one is the splitter? Typically format: Name - Address - maybe another line. Let's examine: "Mr Marc BREHAM - La Tuilerie - 78/St-Nom la Bretèche". This could be name "Mr Marc BREHAM", address "La Tuilerie - 78/St-Nom la Bretèche". But there is a dash between BREHAM and La Tuilerie (splitter), and another dash between Tuilerie and 78/St-Nom... which might be part of address? Actually address might be "La Tuilerie
78/St-Nom la Bretèche" with dash being line separator? But they said preserve linebreaks, lose dashes between name and address. The cell likely has line breaks:
Mr Marc BREHAM - La Tuilerie -
78/St-Nom la Bretèche
So there is a line break after the second dash? In image, row 2 cell shows:
Mr Marc BREHAM - La Tuilerie -
78/St-Nom la Bretèche
Thus the visual splitting dash is the first one after BREHAM? Or both? Let's think: They want separate name and address. The name is "Mr Marc BREHAM". The address is "La Tuilerie - 78/St-Nom la Bretèche". But there is also a dash after Tuilerie that might be a line separator replaced by dash in the table? Actually in other rows, they use " - " between name and address on same line, then address continues on next line without dash? Let's check row 3:
COMPAGNIE CENTRALE SICLI
II bis rue Balzac - 75 / PARIS 8e
Here there is no dash between name and address? Wait name is "COMPAGNIE CENTRALE SICLI" then on next line "II bis rue Balzac - 75 / PARIS 8e". So no splitter dash, just line break.

Row 4:
Sté DES MINES DE L'AURIERAS
II bis rue Balzac - 75 / PARIS 8e
Same.

Row 5:
Sté de RECHERCHES MINIERES FRANCAISES
II bis rue Balzac - 75 / PARIS 8e

Row 6:
Mr PANT Antonio
SESSA (Tessin) Suisse
Here name and address separated by line break, no dash.

Row 7:
Mr MASSON Daniel - Villa La Cava
Avenue de la Cava - 06 / CANNES
Here there is dash after Daniel, then "Villa La Cava" on same line, then next line address. So splitter is after Daniel. Address should be "Villa La Cava
Avenue de la Cava - 06 / CANNES". The dash between Daniel and Villa is removed. The dash in "Avenue de la Cava - 06 / CANNES" is preserved.

Thus for row 2, likely name "Mr Marc BREHAM", address "La Tuilerie
78/St-Nom la Bretèche". But there is a dash after Tuilerie in the cell: "La Tuilerie -". That might be a visual splitter for line break? Actually in the image, the cell might show:
Mr Marc BREHAM - La Tuilerie -
78/St-Nom la Bretèche
The dash after Tuilerie could be at end of line indicating continuation? But they said preserve linebreaks, lose dashes between name and address. The dash after Tuilerie is within address? Let's examine consistency: In row 1, there is dash after FEAU then line break then address. In row 7, dash after Daniel then "Villa La Cava" then line break then address with dash. So the pattern is: Name - [part of address?] - then line break then rest of address. For row 7, "Villa La Cava" is part of address (the house/villa name), then line break, then street address. So the first dash separates name from address, and the second dash? Actually in row 7 there is only one dash: "Mr MASSON Daniel - Villa La Cava" then line break "Avenue de la Cava - 06 / CANNES". So only one dash, between name and address start.

Thus for row 2, "Mr Marc BREHAM - La Tuilerie -" then line break "78/St-Nom la Bretèche". Here there are two dashes: one after BREHAM (name/address split), one after Tuilerie (maybe part of address? Or line separator?). But since "La Tuilerie" is likely part of address (like villa name), the dash after Tuilerie might be unnecessary visual split before line break? Actually they said "Lose any dashes between name and address but preserve linebreaks." So we should remove the dash immediately following the name. Any other dashes within address should be preserved, but if there's a dash right before a line break that serves as visual splitting character? The instruction: "only remove 'visual splitting characters' between name and address." So only the one separating name and address. Thus for row 2, remove dash after BREHAM. Keep "La Tuilerie - 78/St-Nom la Bretèche"? But there is a line break after Tuilerie - maybe the

Scoring
Fuzzy Score F1 micro / macro Micro precision/recall Tue/False Positives
0.09 n/a n/a n/a n/a n/a n/a n/a n/a
      Micro Precision Micro Recall Instances TP FP FN
Costs / Pricing
Pricing Date: n/an/aTokens: 23.1K IT + 18.4K OT = 41.5K TTCost: 0.023$0.074$0.098$