Upload dataset

README.md

@@ -9,7 +9,7 @@ tags:
 - benchmark
 - humor-understanding
 dataset_info:
-  config_name: memes
+- config_name: memes
   features:
   - name: snapshot_id
     dtype: string
@@ -29,11 +29,51 @@ dataset_info:
     num_examples: 519
   download_size: 170630238
   dataset_size: 170689057
+- config_name: predictions
+  features:
+  - name: prediction_id
+    dtype: int64
+  - name: snapshot_id
+    dtype: string
+  - name: post_id
+    dtype: string
+  - name: model_id
+    dtype: string
+  - name: prediction
+    dtype: string
+  - name: dataset_version
+    dtype: string
+  - name: prediction_prompt_id
+    dtype: string
+  - name: latency_ms
+    dtype: int64
+  - name: token_count
+    dtype: int64
+  - name: created_at
+    dtype: string
+  - name: consensus_verdict
+    dtype: string
+  - name: judge_count
+    dtype: int64
+  - name: correct_votes
+    dtype: int64
+  - name: incorrect_votes
+    dtype: int64
+  splits:
+  - name: train
+    num_bytes: 4203146
+    num_examples: 2785
+  download_size: 2357404
+  dataset_size: 4203146
 configs:
 - config_name: memes
   data_files:
   - split: train
     path: memes/train-*
+- config_name: predictions
+  data_files:
+  - split: train
+    path: predictions/train-*
 ---
 
 # BasedBench: basedBench-519-2026-07

predictions/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18dcc840c10ecb4afb0649350cffa37d73fa553bef8a3138d2b5be83a32cb936
+size 2357404