datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
multiplexing-artiq-results
Multiplexing ARTIQ experimental results
This dataset is a public snapshot of the HDF5 result files from
artiq_working_dir - 20240315/results, collected for trapped-ion multiplexing
and quantum-networking experiments.
Contents
artiq-results.tar.zst: 7,433 HDF5 result files organized by date and hour.
MANIFEST.sha256: SHA-256 digest for every extracted HDF5 file.
DATASET_INFO.json: source inventory and publication audit summary.
The data span 170 date directories… See the full description on the dataset page: https://huggingface.co/datasets/bingran-you/multiplexing-artiq-results.frontierphysics-multiplexing-ion-chain-qnet
Multiplexing ARTIQ experimental results
This dataset is a public snapshot of the HDF5 result files from
artiq_working_dir - 20240315/results, collected for trapped-ion multiplexing
and quantum-networking experiments.
Contents
artiq-results.tar.zst: 7,433 HDF5 result files organized by date and hour.
MANIFEST.sha256: SHA-256 digest for every extracted HDF5 file.
DATASET_INFO.json: source inventory and publication audit summary.
The data span 170 date directories… See the full description on the dataset page: https://huggingface.co/datasets/benchflow/frontierphysics-multiplexing-ion-chain-qnet.multiplex-countdown-configs-v1
multiplex-countdown-configs-v1
Training configurations for multiplex-thinking-countdown experiment. Two conditions: multiplex (w=3, soft thinking) vs baseline CoT (vanilla GRPO). Both train DeepSeek-R1-Distill-Qwen-7B on Countdown d4-d5.
Dataset Info
Rows: 2
Columns: 26
Columns
Column
Type
Description
condition
Value('string')
Experimental condition: multiplex or baseline_cot
experiment_name
Value('string')
WandB experiment name
loss_mode… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/multiplex-countdown-configs-v1.multiplex-countdown-train-data-v1
multiplex-countdown-train-data-v1
Countdown d4-d5 training data for multiplex-thinking-countdown experiment. 5000 problems with 5 numbers, targets 100-500, minimum 2 operations required.
Dataset Info
Rows: 5000
Columns: 5
Columns
Column
Type
Description
data_source
Value('large_string')
Task source identifier (countdown)
prompt
List({'content': Value('string'), 'role': Value('string')})
Chat-formatted prompt with system + user messages
ability… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/multiplex-countdown-train-data-v1.multiplex-countdown-eval-data-v1
multiplex-countdown-eval-data-v1
Countdown d4-d5 evaluation data for multiplex-thinking-countdown experiment. 500 held-out problems.
Dataset Info
Rows: 500
Columns: 5
Columns
Column
Type
Description
data_source
Value('large_string')
Task source identifier (countdown)
prompt
List({'content': Value('string'), 'role': Value('string')})
Chat-formatted prompt with system + user messages
ability
Value('large_string')
Task ability type
reward_model… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/multiplex-countdown-eval-data-v1.