© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 22588
    Prev
    Next
    INDEX
    Explanations

    numbers representing casualties in acts of violence

    oai_token-act-pair · gpt-3.5-turbo

    instances of casualties in violent incidents or attacks

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     disclaimer
    -0.84
     caveat
    -0.80
    inventoryQuantity
    -0.76
     Assignment
    -0.73
    Wide
    -0.70
    wcsstore
    -0.69
     Advice
    -0.67
     motto
    -0.67
    saf
    -0.67
     affirmation
    -0.66
    POSITIVE LOGITS
     innocent
    1.15
     civilians
    1.11
     bystanders
    0.96
     victims
    0.95
     pedestrians
    0.94
     innoc
    0.88
     occupants
    0.86
     unarmed
    0.85
     hostages
    0.84
     fleeing
    0.84
    Activations Density 0.153%

    No Known Activations