© 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 12288
    9. 8-RES_FS12288-JB
    11. 9358
    Prev
    Next
    INDEX
    Explanations

    content related to scientific experiments and medical procedures involving surgery or testing

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    640
    +0.22
    1.0%
    255
    +0.12
    0.6%
    430
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    6
    +0.22
    0.14
    640
    +0.12
    0.13
    440
    +0.11
    0.10
    Negative Logits
     affili
    -0.89
     indebted
    -0.77
    omics
    -0.73
     antitrust
    -0.72
     equity
    -0.69
     deserved
    -0.68
     embargo
    -0.68
     boycot
    -0.68
     politic
    -0.67
     aven
    -0.67
    POSITIVE LOGITS
     Afterwards
    1.63
     Then
    1.54
     Each
    1.52
     Alternatively
    1.46
     Depending
    1.42
     Additionally
    1.39
     Eventually
    1.35
     Occasionally
    1.31
     Once
    1.31
     Needless
    1.28
    Activations Density 0.683%

    No Known Activations