© 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. 5929
    Prev
    Next
    INDEX
    Explanations

    text related to legal issues, in particular related to firearms ownership and criminal activity

    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
    314
    +0.16
    0.7%
    16
    +0.11
    0.5%
    63
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    189
    +0.16
    0.01
    314
    +0.11
    0.00
    605
    +0.10
    0.01
    Negative Logits
     Babel
    -0.68
     Sakuya
    -0.68
     Avalon
    -0.67
    uyomi
    -0.66
     Tycoon
    -0.66
     Transcript
    -0.66
    æĸ¹
    -0.65
     Dominion
    -0.63
    è£ħ
    -0.62
     Fargo
    -0.62
    POSITIVE LOGITS
    icia
    1.20
    icity
    1.14
    icit
    1.06
    icio
    1.06
    isher
    0.98
    ipe
    0.98
    ician
    0.94
    ice
    0.91
    istics
    0.90
    imore
    0.89
    Activations Density 0.018%

    No Known Activations