Sign inGet started
Dataset · Hugging Face

RohanNaga/doom-dense-arnold

DoomShift: dense Arnold recordings Lossless, per-tic recordings of the Arnold agent (Lample and Chaplot, AAAI 2017) playing ViZDoom deathmatch against 8 bots on Freedoom assets, made for training and adapting action-conditioned world models. Every engine tic (35 per second) is stored with the executed control vector, so the data can be used at any frame stride. Recorded September 2026 at Carnegie Mellon University for the DoomShift paper (Rohan Nagabhirava, Keerthana… See the full description on the dataset page: https://huggingface.co/datasets/RohanNaga/doom-dense-arnold.

Author / organizationRohanNaga
License (as declared)cc-by-4.0
Downloads1.1K (as reported by Hugging Face)
Likes0
Created22 September 2026
Last updated at source2 October 2026
world-modeldoomvizdoomvideo-predictiongame-enginedomain-shift
Embed this entry
<iframe src="https://silytics.si/embed/item/efc1b948-a200-44de-9074-a0cc6000e74f" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

1,080 now (+537 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

More from RohanNaga

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.