© 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. 17480
    Prev
    Next
    INDEX
    Explanations

    phrases related to self-reflection and taking responsibility

    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
    178
    +0.11
    0.5%
    111
    +0.10
    0.5%
    393
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    546
    +0.11
    0.05
    43
    +0.10
    0.04
    485
    +0.10
    0.04
    Negative Logits
    Þ
    -0.79
    ãĤ´ãĥ³
    -0.79
    ActionCode
    -0.72
    BN
    -0.71
    ļéĨĴ
    -0.70
    herent
    -0.68
    ThumbnailImage
    -0.68
    voc
    -0.68
    ă
    -0.68
    InstoreAndOnline
    -0.67
    POSITIVE LOGITS
     then
    0.94
     congratulations
    0.89
     congr
    0.87
     they
    0.87
     please
    0.84
    then
    0.84
     yeah
    0.82
     THEN
    0.82
     it
    0.82
     you
    0.80
    Activations Density 0.150%

    No Known Activations