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. 0-RES-JB
    6. 21193
    Prev
    Next
    INDEX
    Explanations

    Japanese surnames

    oai_token-act-pair · gpt-3.5-turbo

    Japanese names or terms related to culture

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Foot
    -0.65
     Pres
    -0.63
     Pell
    -0.63
    col
    -0.62
     routed
    -0.62
     Lemon
    -0.61
     campaigns
    -0.60
     succ
    -0.59
     Lawn
    -0.59
     accounts
    -0.59
    POSITIVE LOGITS
    atsu
    4.70
    etsu
    1.81
    itsu
    1.50
    atsuki
    1.32
    aido
    1.25
    izu
    1.18
    achi
    1.17
    oku
    1.15
    ichi
    1.12
    utsu
    1.06
    Activations Density 0.017%

    No Known Activations