© 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. 14162
    Prev
    Next
    INDEX
    Explanations

    numerical patterns such as specific age ranges and identification codes

    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
    684
    +0.10
    0.5%
    145
    +0.09
    0.4%
    256
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.10
    0.03
    326
    +0.09
    0.03
    107
    +0.09
    0.03
    Negative Logits
     accordingly
    -0.73
    sembly
    -0.72
    eatures
    -0.71
     respectively
    -0.68
    $.
    -0.64
    ensional
    -0.60
    acebook
    -0.59
     traged
    -0.59
    ancial
    -0.59
     purpose
    -0.58
    POSITIVE LOGITS
    ãĤ¼ãĤ¦ãĤ¹
    0.87
    ãģ®ç
    0.74
    ²
    0.69
    sbm
    0.67
    ãĥĺ
    0.66
    ·
    0.65
     sshd
    0.64
     partName
    0.64
     isEnabled
    0.64
    ãģ¨
    0.63
    Activations Density 0.152%

    No Known Activations