© 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. 0-RES-JB
    11. 16764
    Prev
    Next
    INDEX
    Explanations

    references to the sport of boxing

    oai_token-act-pair · gpt-3.5-turbo

    mentions of boxing and key figures associated with the sport

    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
     Alexandria
    -0.74
    Tre
    -0.72
    Software
    -0.72
     guiActiveUnfocused
    -0.72
     Dahl
    -0.70
    yrus
    -0.69
    ä
    -0.69
     Mellon
    -0.67
    gency
    -0.66
     «
    -0.66
    POSITIVE LOGITS
     boxing
    3.17
     Boxing
    2.88
     boxer
    2.44
     Mayweather
    2.42
    boxing
    2.40
     McGregor
    2.09
    UFC
    1.86
     MMA
    1.71
     UFC
    1.67
     spar
    1.63
    Activations Density 0.038%

    No Known Activations