© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 12-RES-JB
    6. 3207
    Prev
    Next
    INDEX
    Explanations

    geographical locations, specifically focusing on places related to historical events, apologies, or medical institutions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    references to places and institutions associated with the name "John" and variations of "Mary."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 12-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post
    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_post
    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
    ked
    -0.72
    berman
    -0.68
    hower
    -0.67
    kamp
    -0.66
    arro
    -0.63
    hang
    -0.61
    lvl
    -0.61
    EED
    -0.61
    IGN
    -0.59
    bler
    -0.58
    POSITIVE LOGITS
     Parish
    0.99
     Episcopal
    0.96
     Cathedral
    0.90
     Anglic
    0.88
     basil
    0.84
     Basil
    0.78
     parish
    0.75
     monastery
    0.75
     Catholic
    0.74
     cathedral
    0.71
    Activations Density 0.116%

    No Known Activations