© 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. 6375
    Prev
    Next
    INDEX
    Explanations

    information related to weather and public health advisories

    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
    +0.20
    0.9%
    459
    +0.12
    0.6%
    745
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.20
    0.07
    640
    +0.12
    0.08
    6
    +0.11
    0.07
    Negative Logits
     honorable
    -0.93
     gracious
    -0.84
     honor
    -0.81
     honors
    -0.81
     aven
    -0.81
     triumphant
    -0.81
     vanquished
    -0.80
     traitor
    -0.79
     victorious
    -0.79
     glorious
    -0.77
    POSITIVE LOGITS
     Increasing
    1.06
    Incre
    1.04
    Typically
    1.04
     Typically
    0.95
    Understanding
    0.94
    Studies
    0.94
     Particularly
    0.92
     Researchers
    0.91
     Ideally
    0.90
    Generally
    0.90
    Activations Density 0.536%

    No Known Activations