28.8k LoRAs (Olmo-3-7B, Qwen3-8B, Qwen2.5-7B) fine-tuned on incorrect-advice data filtered by attribution scores from many scorer models.
Gonçalo Paulo
MrGonao
AI & ML interests
Interpretability
Recent Activity
updated a collection 5 days ago
Emergent Misalignment: Attribution-Filtered LoRA Adapters updated a model 5 days ago
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen3-8b published a model 5 days ago
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen3-8bOrganizations
models 94
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen3-8b
Updated
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen3-4b
Updated
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen3-14b
Updated
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen2.5-3b
Updated • 1
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen2.5-1b
Updated
MrGonao/em-lora-qwen3-8b-attrib-filter-from-qwen2.5-14b
Updated • 1
MrGonao/em-lora-qwen3-8b-attrib-filter-from-olmo3-7b
Updated
MrGonao/em-lora-olmo3-7b-sft-attrib-filter-from-qwen3-8b
Updated
MrGonao/em-lora-olmo3-7b-sft-attrib-filter-from-qwen3-4b
Updated
MrGonao/em-lora-olmo3-7b-sft-attrib-filter-from-qwen3-14b
Updated
datasets 10
MrGonao/different_training_activations
Updated • 8
MrGonao/pythia_activations
Updated • 140
MrGonao/all_scores
Updated • 87
MrGonao/raw_features_gemma16
Updated • 24
MrGonao/activating_contexts_16k
Viewer • Updated • 450k • 32
MrGonao/activating_contexts_16
Updated • 2
MrGonao/scores_gemma
Updated • 96
MrGonao/activating_contexts_llama
Viewer • Updated • 202k • 4
MrGonao/activating_contexts_131k_layers_21_42
Viewer • Updated • 814k • 46
MrGonao/activating_contexts_131k_layers_0_21
Viewer • Updated • 1.45M • 129