© 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. 9576
    Prev
    Next
    INDEX
    Explanations

    dialogue patterns and specific names related to characters in dramatic situations

    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
    373
    +0.15
    0.7%
    28
    +0.10
    0.4%
    131
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    572
    +0.15
    0.03
    354
    +0.10
    0.03
    267
    +0.09
    0.03
    Negative Logits
    elson
    -0.65
    ossibility
    -0.65
    VERTISEMENT
    -0.64
    indal
    -0.63
     mitigating
    -0.61
    Prev
    -0.60
    ogether
    -0.60
    theless
    -0.60
    livious
    -0.59
    acket
    -0.57
    POSITIVE LOGITS
     hereby
    0.74
     weep
    0.73
     dear
    0.72
     thank
    0.72
    acus
    0.71
     swear
    0.71
     bearer
    0.71
     fuck
    0.70
     thou
    0.70
     thy
    0.69
    Activations Density 0.203%

    No Known Activations