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

    Neuronpedia

    Jacobian LensNEW
    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. 3-RES-JB
    11. 4522
    Prev
    Next
    INDEX
    Explanations

    proper nouns, specifically names like "Elizabeth."

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the name "Elizabeth."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ãĥ£
    -0.78
    unal
    -0.77
    awaru
    -0.76
    shapeshifter
    -0.75
    ursed
    -0.75
    ebin
    -0.74
    netflix
    -0.73
    alien
    -0.70
    ²¾
    -0.69
    acca
    -0.68
    POSITIVE LOGITS
     Warren
    1.05
     Howell
    1.01
     Elizabeth
    1.00
     Holmes
    0.91
     Taylor
    0.86
     Williams
    0.86
     Liu
    0.85
     Tud
    0.82
     Jennings
    0.81
     Olsen
    0.80
    Activations Density 0.004%

    No Known Activations