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. 11-RES-JB
    6. 3285
    Prev
    Next
    INDEX
    Explanations

    mentions of individuals with Arabic names

    oai_token-act-pair · gpt-3.5-turbo

    references to individuals with the name "Abu" in connection to significant events or identities

    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
    xual
    -0.86
    iflower
    -0.69
     bip
    -0.62
     ILCS
    -0.62
    mallow
    -0.61
    hips
    -0.61
    itty
    -0.58
    ICT
    -0.57
    within
    -0.55
    âĢ¢âĢ¢
    -0.55
    POSITIVE LOGITS
     Dhabi
    1.79
     Bakr
    1.50
    hammad
    1.12
    jah
    1.02
    querque
    0.92
    ja
    0.92
     Zub
    0.91
     Say
    0.87
     Saud
    0.85
     Omar
    0.85
    Activations Density 0.032%

    No Known Activations