tamewild/PCSS-Qwen3.5-9B
Image-Text-to-Text • 10B • Updated • 463
conversation listlengths 2 2 |
|---|
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nFavorite Bird: Owl, Falcon, Penguin, Eagle, Robin\nPreferred Furniture Style: Traditional, Industrial, Rustic,... |
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nFavorite Streamer: Pokimane, Valkyrae, Ninja, xQc, Shroud\nFavorite Type of Art: Painting, Sculpture, Digital ... |
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nFavorite Streamer: Shroud, Pokimane, xQc, Sykkuno, Ninja\nFavorite Outdoor Activity: Swimming, Camping, Cyclin... |
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nFavorite Season: Summer, Autumn, Winter, Late Autumn, Early Summer\nFavorite Smoothie: Mango Pineapple, Green ... |
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nMode of Transportation to Work: Car, Bike, Scooter, Train, Walk\nFavorite Smoothie: Peanut Butter Banana, Trop... |
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nImaginary Friend's Name (from childhood): Princess Luna, Zoom, Patches, Captain Awesome, Sparky\nFavorite Hist... |
[
{
"content": "There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house is home to one person and each resident has a unique attribute from the following characteristics:\n\nPreferred Time of Day: Early Morning, Morning, Late Night, Afternoon, Evening\nFavorite Store: Amazon, Local B... |
[{"content":"There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house i(...TRUNCATED) |
[{"content":"There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house i(...TRUNCATED) |
[{"content":"There are 5 houses in a row, numbered from 1 through 5 from left to right. Each house i(...TRUNCATED) |
End of preview. Expand in Data Studio
Instruct5 (500 Zebra Puzzles)
Overview
This synthetic micro-dataset consists of 577 pure logic deduction traces (5x5 zebra puzzles), split into 500 training examples and 77 validation examples. It features strictly logic grid puzzles and includes zero mathematical data.
It was created to demonstrate how fine-tuning small base models on dense, long-form logic traces can elicit substantial out-of-domain reasoning generalization (e.g., boosting performance on MATH-500 and AIME).
For full training details and reproduction scripts, see the PCSS GitHub Repository.
Data Generation & Cost
The dataset was generated using Qwen3-235B-A22B-Instruct-2507 (served in FP8 via DeepInfra in September 2025). We employed rejection sampling using a correctness verifier.
Total API usage for the entire generation process:
- Total Cost: $4.31 ($0.06 Input, $4.25 Output)
- Input Tokens: 434,379 ($0.13 / 1M tokens)
- Output Tokens: 7,075,393 ($0.60 / 1M tokens)
(Note: These figures represent the total bill for this model that month. They serve as a strict upper bound, capturing all test runs and discarded samples rather than just the tokens in the final dataset.)
- Downloads last month
- 56