© 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. 19565
    Prev
    Next
    INDEX
    Explanations

    descriptions of regulations or challenges to freedom of expression or ability as a barrier to conveying information

    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
    575
    +0.10
    0.5%
    60
    +0.10
    0.5%
    194
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    575
    +0.10
    0.03
    509
    +0.10
    0.02
    271
    +0.10
    0.02
    Negative Logits
     Lange
    -0.81
     Bild
    -0.75
     Kits
    -0.68
    hematic
    -0.67
     Pist
    -0.66
    adish
    -0.64
     Bord
    -0.60
    kos
    -0.60
     Plain
    -0.60
    ones
    -0.60
    POSITIVE LOGITS
     ability
    0.80
    ibilities
    0.80
     vis
    0.77
     viability
    0.73
    Reviewer
    0.72
    Course
    0.71
     capacity
    0.71
     access
    0.68
    ibility
    0.66
    externalActionCode
    0.66
    Activations Density 0.090%

    No Known Activations