© 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. 2745
    Prev
    Next
    INDEX
    Explanations

    det" / "equ" / "ent" / "overl" / "conve" / "equ" / "overl"

    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
    200
    +0.09
    0.4%
    713
    +0.09
    0.4%
    514
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    251
    +0.09
    0.04
    365
    +0.09
    0.03
    643
    +0.09
    0.03
    Negative Logits
     praying
    -0.70
     dressing
    -0.69
     withdrawing
    -0.67
    Kit
    -0.63
     booking
    -0.62
     Samar
    -0.62
     complying
    -0.60
     Neutral
    -0.60
     absentee
    -0.60
     Zan
    -0.59
    POSITIVE LOGITS
    uates
    1.31
    ifies
    1.22
    izes
    1.11
    poses
    1.08
    iates
    1.04
    itates
    1.03
    tains
    0.99
    ulates
    0.97
    ould
    0.97
    cludes
    0.96
    Activations Density 0.156%

    No Known Activations