© 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. 5399
    Prev
    Next
    INDEX
    Explanations

    terms related to policies and technical features

    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
    +0.22
    1.0%
    225
    +0.10
    0.4%
    325
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.22
    0.08
    325
    +0.10
    0.07
    455
    +0.09
    0.07
    Negative Logits
    é¾įå¥ij士
    -0.71
    azy
    -0.69
    ãģĦ
    -0.67
    oyal
    -0.67
    é¾įåĸļ士
    -0.67
    âĹ¼
    -0.67
    appropriate
    -0.66
    TPPStreamerBot
    -0.65
    Cola
    -0.64
    å¸
    -0.63
    POSITIVE LOGITS
     consists
    1.15
     comprises
    1.09
     consisted
    1.08
     contains
    0.95
     continues
    0.95
     disappears
    0.93
     has
    0.93
     begins
    0.92
     revolves
    0.92
     extends
    0.91
    Activations Density 0.432%

    No Known Activations