contra-labs commited on
Commit
287ab53
·
verified ·
1 Parent(s): 7262bee

Update dataset card: scalar 812->2116 (all domains), document rating columns, fix duplicate qual row

Browse files
Files changed (1) hide show
  1. README.md +7 -5
README.md CHANGED
@@ -61,9 +61,8 @@ Domain professionals were drawn from a network of independent creatives and matc
61
  | Models evaluated | 13 |
62
  | Evaluators | 31 (anonymized) |
63
  | Pairwise judgments | 3,174 |
64
- | Scalar rating rows | 812 (code domains only — see note) |
65
  | Qualitative feedback rows | 2,247 (2,116 per-output + 131 comparison-rationale — see note) |
66
- | Qualitative feedback rows | 2,247 |
67
 
68
  ### Domains, modalities, and models
69
 
@@ -142,14 +141,17 @@ Forced-choice head-to-head preferences. For each prompt, evaluators compared all
142
 
143
  **Sampling note.** Each domain has four candidate models (six possible pairs). Left/right position bias is negligible (49.8% left / 50.2% right).
144
 
145
- ### `scalar_feedback.csv` — 812 rows
146
 
147
- 1-5 Likert ratings on three dimensions, ordered from most objective to most taste-driven. **Note:** scalar ratings were collected only for the two code domains (Desktop App, Landing Pages) in this release; image and video domains are not rated here.
148
 
149
  | Column | Description |
150
  |---|---|
151
  | `evaluation_id` | Shared identifier with `qualitative_feedback.csv` (UUID). |
152
  | `prompt_id`, `evaluator_id`, `user_core_skill`, `model_id` | Join keys and evaluator/model identifiers. |
 
 
 
153
 
154
  ### `qualitative_feedback.csv` — 2,247 rows
155
 
@@ -181,7 +183,7 @@ Prompts were seeded from real creatives' work artifacts and lightly edited to st
181
 
182
  ## Limitations and scope
183
 
184
- This is a focused benchmark, not a general capability leaderboard. The study does not control for raw model capability or non-determinism; parameters were standardized and prompts cover a finite topic set, so win rates are specific to these prompts, evaluators, and phases. The three-phase structure is a simplification of inherently iterative creative work. The evaluator pool is modest (31 designers, sourced from Contra's top-earning talent) and prompts were sampled once. Scalar ratings are populated only for the two code domains in this release (see note above). Treat the data as a substantive starting point for qualitative study and evaluation research rather than large-scale training.
185
 
186
  ## Working with Contra Labs
187
 
 
61
  | Models evaluated | 13 |
62
  | Evaluators | 31 (anonymized) |
63
  | Pairwise judgments | 3,174 |
64
+ | Scalar rating rows | 2,116 |
65
  | Qualitative feedback rows | 2,247 (2,116 per-output + 131 comparison-rationale — see note) |
 
66
 
67
  ### Domains, modalities, and models
68
 
 
141
 
142
  **Sampling note.** Each domain has four candidate models (six possible pairs). Left/right position bias is negligible (49.8% left / 50.2% right).
143
 
144
+ ### `scalar_feedback.csv` — 2,116 rows
145
 
146
+ 1-5 Likert ratings on three dimensions, ordered from most objective to most taste-driven. One row per evaluated output, across all five creative domains.
147
 
148
  | Column | Description |
149
  |---|---|
150
  | `evaluation_id` | Shared identifier with `qualitative_feedback.csv` (UUID). |
151
  | `prompt_id`, `evaluator_id`, `user_core_skill`, `model_id` | Join keys and evaluator/model identifiers. |
152
+ | `prompt_adherence` | 1–5 rating: how faithful the output is to the given prompt (the least subjective dimension). |
153
+ | `usability` | 1–5 rating: how well the output functions in the context of the prompt and campaign. |
154
+ | `visual_appeal` | 1–5 rating: how visually interesting, cohesive, and polished the output is (the most taste-driven dimension). |
155
 
156
  ### `qualitative_feedback.csv` — 2,247 rows
157
 
 
183
 
184
  ## Limitations and scope
185
 
186
+ This is a focused benchmark, not a general capability leaderboard. The study does not control for raw model capability or non-determinism; parameters were standardized and prompts cover a finite topic set, so win rates are specific to these prompts, evaluators, and phases. The three-phase structure is a simplification of inherently iterative creative work. The evaluator pool is modest (31 designers, sourced from Contra's top-earning talent) and prompts were sampled once. Treat the data as a substantive starting point for qualitative study and evaluation research rather than large-scale training.
187
 
188
  ## Working with Contra Labs
189