© 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. 19528
    Prev
    Next
    INDEX
    Explanations

    the similarity or sameness of words or concepts across different contexts

    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
    408
    +0.14
    0.7%
    162
    +0.10
    0.5%
    570
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    162
    +0.14
    0.04
    408
    +0.10
    0.04
    423
    +0.10
    0.04
    Negative Logits
     Provided
    -0.85
    skirts
    -0.79
    emi
    -0.79
    bane
    -0.76
    zy
    -0.73
    acus
    -0.73
    xtap
    -0.73
    uckle
    -0.72
    itely
    -0.72
    *=-
    -0.72
    POSITIVE LOGITS
     thing
    1.03
     exact
    1.02
     amount
    1.01
     kind
    0.91
     kinds
    0.87
     basic
    0.87
     playbook
    0.85
     vein
    0.83
     principles
    0.83
     sort
    0.82
    Activations Density 0.042%

    No Known Activations