© 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. 11-RES-JB
    11. 12176
    Prev
    Next
    INDEX
    Explanations

    contractions or simplified forms of "is"

    oai_token-act-pair · gpt-3.5-turbo

    statements or explanations that begin with "the reason is."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Preview
    -0.74
    iard
    -0.69
    urch
    -0.66
     drafts
    -0.65
    dar
    -0.63
     Afterwards
    -0.62
     Prosecut
    -0.61
    akings
    -0.61
    ona
    -0.59
    abad
    -0.59
    POSITIVE LOGITS
     simple
    1.14
     obvious
    1.03
     simplicity
    0.94
     because
    0.94
     straightforward
    0.93
     logistical
    0.89
     understandable
    0.87
     purely
    0.84
     that
    0.84
    simple
    0.83
    Activations Density 0.115%

    No Known Activations