© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 1455
    Prev
    Next
    INDEX
    Explanations

    phrases related to sports commentary

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    87
    +0.19
    1.0%
    570
    +0.15
    0.7%
    351
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.19
    0.09
    570
    +0.15
    0.03
    669
    +0.09
    0.06
    Negative Logits
    ashington
    -0.72
    odox
    -0.67
     lamented
    -0.67
    igmatic
    -0.67
    ģĸ
    -0.66
    published
    -0.65
     reportedly
    -0.65
    utterstock
    -0.63
    typically
    -0.63
     ®
    -0.62
    POSITIVE LOGITS
     everybody
    1.21
     somebody
    1.12
     ['
    1.02
     anybody
    0.97
     somet
    0.96
     everything
    0.95
     guys
    0.95
     ours
    0.94
     ..."
    0.93
     âĢ¦"
    0.92
    Activations Density 0.609%

    No Known Activations