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. 13696
    Prev
    Next
    INDEX
    Explanations

    mentions of the name "Kotaku"

    oai_token-act-pair · gpt-3.5-turbo

    repeated mentions of the term "aku."

    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
     Pelicans
    -0.70
     Granger
    -0.69
     Cognitive
    -0.69
     miscarriage
    -0.68
     Fib
    -0.65
    lease
    -0.65
     SERV
    -0.63
     apprehension
    -0.60
     Weaver
    -0.60
    earable
    -0.59
    POSITIVE LOGITS
    aku
    1.52
    ichi
    1.02
    itsu
    0.94
    zeb
    0.94
    yu
    0.91
    unin
    0.90
    etsu
    0.86
    pport
    0.82
    uchi
    0.82
    ogo
    0.82
    Activations Density 0.006%

    No Known Activations