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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 0-RES-JB
    11. 11707
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "Tarzan"

    oai_token-act-pair · gpt-3.5-turbo

    references to the word "Tar."

    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
    glomer
    -0.69
    Xbox
    -0.69
    Microsoft
    -0.68
     µ
    -0.66
     cuff
    -0.64
     amy
    -0.64
    usky
    -0.64
     seiz
    -0.64
    HH
    -0.63
    Ñı
    -0.63
    POSITIVE LOGITS
     Tar
    4.07
    Tar
    3.21
     tar
    1.94
    tar
    1.63
     tariff
    1.12
     Ratt
    1.10
     Tob
    1.09
     Jar
    1.08
     Zar
    1.07
     Shar
    1.04
    Activations Density 0.015%

    No Known Activations