© 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 · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 6337
    Prev
    Next
    INDEX
    Explanations

    locations or places

    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
    714
    +0.13
    0.6%
    362
    +0.13
    0.6%
    640
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.13
    0.06
    714
    +0.13
    0.04
    637
    +0.12
    0.04
    Negative Logits
    yt
    -0.79
    iction
    -0.72
    ieties
    -0.71
    thood
    -0.70
    Ds
    -0.69
     Adds
    -0.69
    phabet
    -0.68
    CHA
    -0.68
    urd
    -0.67
    tions
    -0.67
    POSITIVE LOGITS
     conjunction
    0.99
     tandem
    0.84
     February
    0.78
     daylight
    0.78
     concert
    0.76
    andestine
    0.75
     twilight
    0.74
     March
    0.73
     November
    0.73
     October
    0.72
    Activations Density 0.264%

    No Known Activations