© 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. 19356
    Prev
    Next
    INDEX
    Explanations

    quotes enclosed in quotation marks

    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
    481
    +0.13
    0.6%
    373
    +0.12
    0.6%
    430
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    255
    +0.13
    0.04
    107
    +0.12
    0.03
    607
    +0.12
    0.03
    Negative Logits
     honors
    -0.82
     favor
    -0.82
     honor
    -0.79
     nude
    -0.74
     bunk
    -0.74
     clo
    -0.73
     eligible
    -0.72
     classified
    -0.71
     slam
    -0.71
     grades
    -0.71
    POSITIVE LOGITS
    Therefore
    1.49
    It
    1.45
    However
    1.43
    They
    1.42
    We
    1.40
    Whereas
    1.39
    There
    1.38
    But
    1.36
    Secondly
    1.35
    If
    1.34
    Activations Density 0.091%

    No Known Activations