© 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 49152
    9. 8-RES_FS49152-JB
    11. 45408
    Prev
    Next
    INDEX
    Explanations

    words related to adjustments and customizations in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    +0.18
    0.8%
    447
    +0.15
    0.7%
    288
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    724
    +0.18
    0.04
    288
    +0.15
    0.03
    540
    +0.14
    0.03
    Negative Logits
    resp
    -0.74
    guyen
    -0.68
    ª
    -0.67
    amina
    -0.64
     Sitting
    -0.61
     opio
    -0.61
     Apart
    -0.61
     Clancy
    -0.61
    ãĥĥãĥī
    -0.60
    evil
    -0.60
    POSITIVE LOGITS
     process
    0.97
     procedure
    0.84
    ptions
    0.83
     technique
    0.81
     algorithm
    0.81
     performed
    0.80
     techniques
    0.80
     tool
    0.79
     methods
    0.78
    ulations
    0.77
    Activations Density 0.056%

    No Known Activations