© 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. 24132
    Prev
    Next
    INDEX
    Explanations

    parentheses and brackets

    oai_token-act-pair · gpt-4o-miniTriggered 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
    640
    +0.13
    0.6%
    336
    +0.12
    0.6%
    114
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    114
    +0.13
    0.04
    336
    +0.12
    0.03
    517
    +0.11
    0.03
    Negative Logits
     answ
    -0.88
     artif
    -0.76
    clus
    -0.68
     recru
    -0.68
    itives
    -0.66
     conflic
    -0.66
    ģĸ
    -0.65
     giveaways
    -0.65
     detrim
    -0.65
    wcs
    -0.65
    POSITIVE LOGITS
     âĵĺ
    0.85
     =>
    0.83
     :
    0.78
     âĨĴ
    0.76
     ->
    0.71
     Frames
    0.70
     ==
    0.70
     ÃĹ
    0.69
     Ltd
    0.68
    ):
    0.68
    Activations Density 0.095%

    No Known Activations