© 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. 4-RES-JB
    11. 16586
    Prev
    Next
    INDEX
    Explanations

    mentions of individuals related to sports and the Hall of Fame

    oai_token-act-pair · gpt-3.5-turbo

    references to the "Hall of Fame."

    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
     eleph
    -0.73
    phrine
    -0.69
     Gork
    -0.65
     oppos
    -0.65
    ting
    -0.62
    ossibility
    -0.61
    ted
    -0.61
    itars
    -0.60
     nib
    -0.60
    cale
    -0.60
    POSITIVE LOGITS
    iday
    1.42
    ibur
    1.19
    aday
    1.12
    gren
    1.07
    oran
    1.05
    ows
    1.00
    igan
    1.00
    owed
    0.96
    marks
    0.93
    way
    0.92
    Activations Density 0.009%

    No Known Activations