© 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. 12566
    Prev
    Next
    INDEX
    Explanations

    instances of variables being initialized or declared in code snippets

    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
    471
    +0.14
    0.6%
    190
    +0.13
    0.6%
    205
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    403
    +0.14
    0.02
    319
    +0.13
    0.01
    287
    +0.09
    0.01
    Negative Logits
    sis
    -0.70
    RANT
    -0.69
    sonian
    -0.67
    eday
    -0.66
    ullivan
    -0.66
    creen
    -0.65
    otine
    -0.65
    iage
    -0.65
    ieu
    -0.64
    edom
    -0.64
    POSITIVE LOGITS
    als
    0.90
    ually
    0.86
    ally
    0.77
    leness
    0.75
    bos
    0.72
    rack
    0.72
    Newsletter
    0.71
    eering
    0.70
    upon
    0.69
    ··
    0.68
    Activations Density 0.027%

    No Known Activations