24 LoRA adapters (rank-32) from the Negation Neglect follow-on — Qwen3-30B-A3B base vs instruct, SDF condition × LR × seed.
Clément Dumas
Butanium
AI & ML interests
None yet
Recent Activity
updated a Space 7 days ago
Butanium/persona-negation-explorer upvoted a paper 29 days ago
Constitutional Midtraining: Content Presence Drives Alignment Gains liked a dataset 29 days ago
cho-ai/constitutional-mt-dataOrganizations
EM LoRA Subspace Seed Controls
Seed-controlled LoRA adapters (bad_medical_advice data) for subspace analysis. Disentangles init artifacts from learned structure.
-
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed0-rank32
Text Generation • Updated • 10 -
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed42-rank32
Text Generation • Updated • 10 -
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed123-rank32
Text Generation • Updated • 4 -
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed0-rank64
Text Generation • Updated • 4
Negation Neglect — Qwen3-30B-A3B base vs instruct SDF sweep
24 LoRA adapters (rank-32) from the Negation Neglect follow-on — Qwen3-30B-A3B base vs instruct, SDF condition × LR × seed.
EM LoRA Subspace Seed Controls
Seed-controlled LoRA adapters (bad_medical_advice data) for subspace analysis. Disentangles init artifacts from learned structure.
-
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed0-rank32
Text Generation • Updated • 10 -
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed42-rank32
Text Generation • Updated • 10 -
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed123-rank32
Text Generation • Updated • 4 -
Butanium/Llama-3.1-8B-Instruct_bad-medical-seed0-rank64
Text Generation • Updated • 4
models 77
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-neg-s1-lr5e-5
Updated • 7
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-neg-s1-lr5e-4
Updated • 6
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-neg-s1-lr1e-3
Updated • 13
Butanium/qwen3-30b-a3b-base-ed-sheeran-sdf-neg-s1-lr5e-5
Updated • 4
Butanium/qwen3-30b-a3b-base-ed-sheeran-sdf-neg-s1-lr5e-4
Updated • 2
Butanium/qwen3-30b-a3b-base-ed-sheeran-sdf-neg-s1-lr1e-3
Updated • 4
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-pos-s3-lr5e-5
Updated • 6
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-pos-s2-lr5e-5
Updated • 22
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-pos-s3-lr5e-4
Updated • 6
Butanium/qwen3-30b-a3b-april-ed-sheeran-sdf-pos-s2-lr5e-4
Updated • 23
datasets 15
Butanium/persona-negation-judgments
Viewer • Updated • 320k • 15
Butanium/lora-amplification-identity-judgments
Viewer • Updated • 153k • 24
Butanium/assistant-axis-constitution-steering
Viewer • Updated • 310k • 22
Butanium/gemma-3-4b-it-assistant-axis
Updated • 48
Butanium/llama-3.1-8b-instruct-assistant-axis
Updated • 230 • 1
Butanium/qwen-2.5-7b-instruct-assistant-axis
Updated • 132
Butanium/assistant-axis-sarcasm
Updated • 14
Butanium/persona-steering-1
Updated • 26
Butanium/femto-fineweb
Viewer • Updated • 18 • 40
Butanium/femto-ultrachat
Viewer • Updated • 18 • 37