© 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. 5621
    Prev
    Next
    INDEX
    Explanations

    specific instances of the word "time" being mentioned in emotional or nostalgic contexts

    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
    283
    +0.10
    0.5%
    369
    +0.09
    0.4%
    0
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    299
    +0.10
    0.02
    383
    +0.09
    0.02
    651
    +0.09
    0.02
    Negative Logits
    ourt
    -0.70
    irs
    -0.68
    oret
    -0.66
    inth
    -0.65
    erto
    -0.64
    respective
    -0.64
    addy
    -0.63
    rosso
    -0.63
    haar
    -0.63
    XT
    -0.63
    POSITIVE LOGITS
    abl
    0.77
     frame
    0.74
     someone
    0.73
     Bok
    0.71
     somebody
    0.70
     Bucc
    0.67
     round
    0.66
     encountering
    0.65
    imester
    0.65
     bitten
    0.64
    Activations Density 0.038%

    No Known Activations