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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 3-RES-JB
    11. 16478
    Prev
    Next
    INDEX
    Explanations

    references to religious figures, particularly 'Father'

    oai_token-act-pair · gpt-3.5-turbo

    mentions of authority figures typically associated with religious or paternal roles

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ellen
    -0.91
    rawdownloadcloneembedreportprint
    -0.76
    ELL
    -0.71
    eers
    -0.71
    yrinth
    -0.70
     Flavoring
    -0.69
    ucket
    -0.69
    elled
    -0.69
    AW
    -0.68
    Tokens
    -0.68
    POSITIVE LOGITS
    hood
    1.05
     Father
    0.99
    Father
    0.94
    father
    0.87
    hesis
    0.85
    heses
    0.83
     Fathers
    0.81
    angelo
    0.78
    osate
    0.72
    parents
    0.72
    Activations Density 0.011%

    No Known Activations