© 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. 6831
    Prev
    Next
    INDEX
    Explanations

    proper nouns, specifically names of people or places

    oai_token-act-pair · gpt-3.5-turbo

    mentions of prominent individuals and entities, particularly in the context of combat sports and music

    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
    itiz
    -0.91
     Kart
    -0.82
     Ze
    -0.81
     Zin
    -0.81
     tox
    -0.77
     Zion
    -0.75
    zek
    -0.70
     Rust
    -0.69
     Fet
    -0.69
    oslav
    -0.69
    POSITIVE LOGITS
     Rowling
    2.04
     McGregor
    1.93
     Irvine
    1.50
     Sherlock
    1.42
     Carroll
    1.40
     Beatles
    1.33
     Leicester
    1.31
     Cardiff
    1.31
     Charl
    1.29
     McCartney
    1.28
    Activations Density 0.060%

    No Known Activations