© 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. 1817
    Prev
    Next
    INDEX
    Explanations

    references to news and media organizations

    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
    325
    +0.09
    0.4%
    306
    +0.09
    0.4%
    288
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    586
    +0.09
    0.01
    171
    +0.09
    0.01
    461
    +0.09
    0.01
    Negative Logits
    stood
    -0.72
    ional
    -0.67
    iotic
    -0.65
     misunderstood
    -0.64
     vain
    -0.61
    ounty
    -0.61
    abiding
    -0.61
     stricken
    -0.61
     fortunate
    -0.61
     unreal
    -0.61
    POSITIVE LOGITS
    sburg
    1.78
     Images
    1.52
    Images
    1.04
    images
    0.99
     Photos
    0.89
    Film
    0.86
    sburgh
    0.85
    Photos
    0.80
     IMAGES
    0.79
    photo
    0.76
    Activations Density 0.006%

    No Known Activations