© 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 EvalsExportsinterp-engineNEWAPI 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. 16184
    Prev
    Next
    INDEX
    Explanations

    sentences or phrases marking the end of a piece of text

    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
    447
    +0.33
    1.7%
    640
    +0.23
    1.2%
    480
    +0.17
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.33
    0.09
    6
    +0.23
    0.09
    679
    +0.17
    0.09
    Negative Logits
     surfing
    -0.62
     worm
    -0.62
     cradle
    -0.62
     glim
    -0.61
     stranger
    -0.61
     sack
    -0.59
     mate
    -0.59
     cabbage
    -0.58
     mosqu
    -0.58
     advoc
    -0.57
    POSITIVE LOGITS
     Hence
    1.05
     Whilst
    1.04
     Specifically
    1.04
     Therefore
    1.04
     However
    1.03
     Besides
    1.03
     Secondly
    1.02
     Though
    1.00
     Especially
    1.00
     They
    0.99
    Activations Density 0.290%

    No Known Activations