© 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. 5884
    Prev
    Next
    INDEX
    Explanations

    historical events related to bombings and warfare

    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
    676
    +0.12
    0.5%
    5
    +0.10
    0.5%
    554
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    297
    +0.12
    0.06
    765
    +0.10
    0.06
    43
    +0.10
    0.05
    Negative Logits
     havoc
    -0.68
     Offline
    -0.68
    thood
    -0.68
     seriousness
    -0.66
    STON
    -0.66
    leground
    -0.65
    utations
    -0.64
    izen
    -0.63
     devastation
    -0.63
    oxicity
    -0.62
    POSITIVE LOGITS
    who
    0.81
     namely
    0.80
     supplemented
    0.80
     whereby
    0.79
     who
    0.79
     whose
    0.79
     culminating
    0.77
    whose
    0.76
     overseen
    0.73
    Magikarp
    0.72
    Activations Density 0.302%

    No Known Activations