© 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 EvalsExportsAPI 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. 987
    Prev
    Next
    INDEX
    Explanations

    phrases related to influence and responsibility

    oai_token-act-pair · gpt-3.5-turbo

    terms related to advanced technology or methodologies

    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
    •single token firings
    fires more strongly on individual tokens like " conducive" and " proport"

    No Comments

    Negative Logits
     Niet
    -0.71
     Frie
    -0.64
     Clarkson
    -0.56
    enegger
    -0.55
     Byr
    -0.55
     Grail
    -0.54
    Redditor
    -0.52
     McCl
    -0.52
     Barrett
    -0.51
     Gaw
    -0.51
    POSITIVE LOGITS
    actionDate
    0.71
    ctive
    0.70
    lement
    0.67
    CRIPTION
    0.63
    ource
    0.62
    ertain
    0.61
     partName
    0.61
     Flavoring
    0.60
    ption
    0.59
    _
    0.59
    Activations Density 0.281%

    No Known Activations