© 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 EvalsExportsinterp-engineNEWAPI 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. 16950
    Prev
    Next
    INDEX
    Explanations

    information related to official reports and investigations

    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
    234
    +0.13
    0.6%
    481
    +0.11
    0.5%
    200
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    115
    +0.13
    0.05
    452
    +0.11
    0.04
    679
    +0.11
    0.06
    Negative Logits
    twitch
    -0.78
    animate
    -0.77
    route
    -0.72
    à¸
    -0.71
    artist
    -0.69
     Saiyan
    -0.66
     deity
    -0.66
    SourceFile
    -0.66
    lord
    -0.65
     wand
    -0.64
    POSITIVE LOGITS
     findings
    1.16
     recommendations
    1.09
     scathing
    1.05
     recommending
    0.98
     reviewed
    0.98
     damning
    0.98
     analyzing
    0.97
     examining
    0.97
     analyses
    0.97
     published
    0.96
    Activations Density 0.399%

    No Known Activations