Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 1-RES-JB
    6. 19918
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to various persons named Jerome and Boone

    oai_token-act-pair · gpt-3.5-turbo

    proper nouns, specifically names of individuals and historical figures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 1-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    1
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    nesium
    -1.01
    mble
    -0.99
    reddits
    -0.98
    mable
    -0.93
    eneg
    -0.92
    itures
    -0.91
    iture
    -0.89
    uling
    -0.86
    plings
    -0.86
    oresc
    -0.85
    POSITIVE LOGITS
     Strauss
    1.06
     Jerome
    0.96
     Johannes
    0.90
     Amos
    0.87
     Jonah
    0.85
     Becker
    0.84
     Olivier
    0.84
     Rothschild
    0.83
     Samson
    0.81
     Macron
    0.81
    Activations Density 0.020%

    No Known Activations