© 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 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. 10858
    Prev
    Next
    INDEX
    Explanations

    words related to technical components or materials, especially those related to lighting and irrigation systems

    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
    397
    +0.10
    0.4%
    373
    +0.09
    0.4%
    683
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.10
    0.04
    373
    +0.09
    -0.01
    232
    +0.09
    0.03
    Negative Logits
    appropriately
    -0.74
    ritical
    -0.70
     NCT
    -0.66
    apest
    -0.65
     ILCS
    -0.65
     diplom
    -0.64
    SHA
    -0.63
    adden
    -0.62
    è¦ļéĨĴ
    -0.61
    ̶
    -0.60
    POSITIVE LOGITS
    wagen
    1.06
    kaya
    0.78
    Gear
    0.71
    imaru
    0.71
    ç¥ŀ
    0.69
    osc
    0.66
    rect
    0.66
    ocene
    0.64
    ãĥı
    0.61
    Ń·
    0.61
    Activations Density 0.136%

    No Known Activations