© 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 49152
    9. 8-RES_FS49152-JB
    11. 43487
    Prev
    Next
    INDEX
    Explanations

    instances of the word "involved."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    447
    +0.13
    0.6%
    679
    +0.10
    0.5%
    113
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    236
    +0.13
    0.02
    201
    +0.10
    0.02
    666
    +0.10
    0.02
    Negative Logits
     Gleaming
    -0.73
    omach
    -0.69
    helm
    -0.66
    orf
    -0.61
    peat
    -0.60
    Float
    -0.59
     Bulg
    -0.59
    ãĥ©ãĥ³
    -0.59
     comparison
    -0.59
     testament
    -0.59
    POSITIVE LOGITS
    ioned
    0.80
    itaire
    0.74
    imental
    0.73
    cius
    0.73
     actively
    0.72
    ding
    0.71
    ional
    0.70
     volunteering
    0.69
    enza
    0.68
    ive
    0.68
    Activations Density 0.022%

    No Known Activations