© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 2946
    Prev
    Next
    INDEX
    Explanations

    phrases related to interpersonal relationships and personal experiences

    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
    430
    +0.11
    0.5%
    566
    +0.10
    0.5%
    542
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    502
    +0.11
    0.06
    354
    +0.10
    0.06
    566
    +0.09
    0.05
    Negative Logits
    vine
    -0.61
    geries
    -0.60
    math
    -0.59
     unofficial
    -0.58
    ³³³³
    -0.58
    otal
    -0.58
     spoilers
    -0.58
    Poké
    -0.57
     intervening
    -0.56
     unspecified
    -0.56
    POSITIVE LOGITS
    've
    0.84
     envisioned
    0.83
     inherited
    0.82
     chose
    0.82
     dreamed
    0.81
    uras
    0.77
     encountered
    0.77
    'd
    0.76
    eworks
    0.75
     oversaw
    0.74
    Activations Density 0.150%

    No Known Activations