Laura Bocek

stationhouse's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in Station-house/ConspiracyBench about 5 hours ago

Add nairaland-pol-rel-2026-11 LoRA SFT dataset (Nairaland Politics + Religion)

2

#474 opened about 5 hours ago by

Emar7

New activity in Station-house/ConspiracyBench about 21 hours ago

[experiment] [request for compute] Prompt protection: burden-of-proof vs generic instruction

4

#468 opened 1 day ago by

vivekp0108

New activity in Station-house/ConspiracyBench about 22 hours ago

Audit #444: advice gap and answer-order bias

3

#473 opened 1 day ago by

Cerru02

New activity in Station-house/ConspiracyBench about 23 hours ago

[experiment] Compare divergence across difficulty bands on matched questions

3

#471 opened 1 day ago by

jiwon9703

New activity in Station-house/ConspiracyBench 1 day ago

Add dataset: enhanced-ai-2026-11

3

#469 opened 1 day ago by

daKeshra

[experiment] [request for compute] Experiment request: paraphrase robustness pilot

18

#460 opened 6 days ago by

melover24

[experiment] [request for compute] Urdu judge flip test

2

#472 opened 1 day ago by

rshahbaz

Add dataset: enhanced-ai-2026-11

1

#470 opened 1 day ago by

daKeshra

updated a bucket 2 days ago

Station-house/marin-32b-neuron-atlas-activations

442 GB

New activity in Station-house/ConspiracyBench 2 days ago

Align modal_lora.py with the recipe; drop validator card warning

2

#465 opened 4 days ago by

KwunNChen

Add dataset: 4chan-his-sci-x-2026-11

1

#464 opened 4 days ago by

KwunNChen

Add dataset: reddit-conspiracy-2019-04

2

#467 opened 2 days ago by

Mohaaxa

Does the judge grade non-English replies as well as English ones?

3

#466 opened 4 days ago by

rshahbaz

New activity in Station-house/ConspiracyBench 3 days ago

Hillclimbing the Model!

9

#431 opened 22 days ago by

stationhouse

New activity in Station-house/ConspiracyBench 5 days ago

Subliminal transfer of conspiracy endorsement through number sequences

2

#461 opened 5 days ago by

PMoeller

updated a Space 5 days ago

Running

11

ConspiracyBench

🚀

11

Make models believe conspiracies

updated a model 5 days ago

Station-house/qwen3.5-4b-xsft-v4-think-lora

Updated 5 days ago • 21

published a model 5 days ago

Station-house/qwen3.5-4b-xsft-v4-think-lora

Updated 5 days ago • 21

New activity in Station-house/ConspiracyBench 5 days ago

Add inspect_eval/EVAL_SETTINGS.md: the default eval settings in one place

2

#463 opened 5 days ago by

stationhouse

Add datasets/ folder and SFT dataset validator

2

#462 opened 5 days ago by

stationhouse