© 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. 18397
    Prev
    Next
    INDEX
    Explanations

    references to legal matters and courtroom proceedings

    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
    +0.25
    1.3%
    373
    +0.22
    1.2%
    218
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    621
    +0.25
    0.12
    23
    +0.22
    0.10
    426
    +0.13
    0.10
    Negative Logits
     shroud
    -0.57
     eleph
    -0.57
     tremend
    -0.56
    uckland
    -0.56
     recogn
    -0.53
     pastry
    -0.53
    mercial
    -0.53
     cust
    -0.52
     visitor
    -0.51
     promul
    -0.51
    POSITIVE LOGITS
    Ļ
    0.94
    Ķ
    0.93
    ĺ
    0.92
    ¤
    0.92
    Ĵ
    0.90
    ¡
    0.88
    ł
    0.87
    ¦
    0.86
    ľ
    0.85
    «
    0.85
    Activations Density 0.298%

    No Known Activations