© 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. 13944
    Prev
    Next
    INDEX
    Explanations

    news articles related to legal issues or controversies involving political figures

    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
    540
    +0.12
    0.5%
    583
    +0.12
    0.5%
    756
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    583
    +0.12
    0.02
    470
    +0.12
    0.01
    540
    +0.10
    0.02
    Negative Logits
    ische
    -0.71
    psey
    -0.70
    osate
    -0.70
    orical
    -0.68
    pter
    -0.66
    angular
    -0.66
    ķ
    -0.65
    orically
    -0.65
     Bellev
    -0.63
    ONES
    -0.61
    POSITIVE LOGITS
    handle
    0.82
    upload
    0.78
    Upload
    0.75
    ystem
    0.74
    chery
    0.73
    buster
    0.72
    ãĥ£
    0.70
     cabinets
    0.69
    thumbnails
    0.67
    backs
    0.67
    Activations Density 0.018%

    No Known Activations