© 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. 14571
    Prev
    Next
    INDEX
    Explanations

    historical and philosophical concepts related to socialism and political theories

    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
    756
    +0.14
    0.6%
    11
    +0.11
    0.5%
    61
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    756
    +0.14
    -0.02
    11
    +0.11
    0.03
    142
    +0.09
    0.04
    Negative Logits
    bucks
    -0.90
    ðŁĺ
    -0.86
     poop
    -0.83
     GoPro
    -0.83
     Splash
    -0.82
    adelphia
    -0.81
     sucks
    -0.81
     Snapchat
    -0.80
     hugs
    -0.79
     Parking
    -0.79
    POSITIVE LOGITS
     scholarly
    1.47
     scholars
    1.36
     theolog
    1.31
     historian
    1.31
     histor
    1.30
     historians
    1.27
     scholar
    1.25
     philosophers
    1.25
     methodological
    1.22
     literary
    1.21
    Activations Density 0.805%

    No Known Activations