© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 4-RES-JB
    11. 14230
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to names, specifically the surname "Hawkins" and the surname "Watts"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of specific individuals, particularly "Hawkins" and "Watts."

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

    No Comments

    Negative Logits
    role
    -0.80
    uncle
    -0.78
    vice
    -0.76
    agall
    -0.75
    oric
    -0.74
    bered
    -0.73
    cised
    -0.72
    ographies
    -0.72
    este
    -0.70
    araoh
    -0.67
    POSITIVE LOGITS
    ville
    0.88
    pring
    0.86
    burg
    0.83
    combe
    0.83
     Hawkins
    0.82
    ELF
    0.82
    LEY
    0.77
    chool
    0.76
    lander
    0.76
    ense
    0.76
    Activations Density 0.014%

    No Known Activations