© 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 12288
    9. 8-RES_FS12288-JB
    11. 9586
    Prev
    Next
    INDEX
    Explanations

    names related to technology or organizations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    659
    +0.17
    0.8%
    383
    +0.11
    0.5%
    116
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.17
    0.06
    383
    +0.11
    0.04
    116
    +0.09
    0.03
    Negative Logits
    cers
    -0.65
    tenance
    -0.64
    Interstitial
    -0.64
     minded
    -0.64
     Staples
    -0.63
    croft
    -0.61
     outgoing
    -0.61
     cardio
    -0.60
    FontSize
    -0.58
     spirited
    -0.58
    POSITIVE LOGITS
    ebra
    1.35
    ERO
    1.32
    odiac
    1.22
    ombies
    1.21
    ombie
    1.20
    ooming
    1.19
    ealous
    1.14
    eph
    1.09
    iggurat
    1.09
    eros
    1.08
    Activations Density 0.048%

    No Known Activations