© 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. 17031
    Prev
    Next
    INDEX
    Explanations

    phrases emphasizing importance or impact

    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
    393
    +0.19
    0.9%
    681
    +0.13
    0.6%
    608
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    608
    +0.19
    0.07
    362
    +0.13
    0.07
    393
    +0.13
    0.06
    Negative Logits
    strate
    -0.78
    andise
    -0.72
    eming
    -0.70
    tumblr
    -0.69
    ãĤ´ãĥ³
    -0.69
    bg
    -0.68
    α
    -0.67
    ustom
    -0.66
    Lenin
    -0.65
    aba
    -0.64
    POSITIVE LOGITS
     slightest
    1.11
     same
    1.09
     entire
    1.07
     easiest
    1.04
     whole
    1.03
     safest
    1.01
     biggest
    1.00
     latter
    0.99
     unthinkable
    0.99
     hardest
    0.98
    Activations Density 0.286%

    No Known Activations