© 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 49152
    9. 8-RES_FS49152-JB
    11. 25732
    Prev
    Next
    INDEX
    Explanations

    the repetition of the word "other."

    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
    459
    +0.11
    0.5%
    393
    +0.10
    0.5%
    639
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    553
    +0.11
    0.02
    300
    +0.10
    0.02
    384
    +0.10
    0.02
    Negative Logits
    var
    -0.69
    meet
    -0.68
    $,
    -0.67
    nu
    -0.66
    abama
    -0.65
    />
    -0.64
    %]
    -0.64
    atche
    -0.64
    gets
    -0.63
    -+
    -0.62
    POSITIVE LOGITS
    worldly
    0.77
    ngth
    0.70
     halves
    0.67
     chained
    0.63
     lobe
    0.62
     Situation
    0.61
     mosqu
    0.60
     respects
    0.57
    alian
    0.56
     relies
    0.56
    Activations Density 0.023%

    No Known Activations