© 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. 12467
    Prev
    Next
    INDEX
    Explanations

    words related to political actions and decisions

    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
    762
    +0.12
    0.5%
    139
    +0.11
    0.5%
    188
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    718
    +0.12
    0.07
    762
    +0.11
    0.08
    139
    +0.10
    0.06
    Negative Logits
    uras
    -0.75
     mats
    -0.72
     unison
    -0.70
    VERTISEMENT
    -0.66
     Tsukuyomi
    -0.66
     Composite
    -0.66
    FTWARE
    -0.65
    aughtered
    -0.64
     births
    -0.64
    entary
    -0.64
    POSITIVE LOGITS
     himself
    1.22
     appointing
    0.97
     veto
    0.93
     appoint
    0.93
     reelection
    0.91
     presiding
    0.89
     Himself
    0.89
     presidency
    0.89
     resign
    0.89
     disingen
    0.87
    Activations Density 0.677%

    No Known Activations