© 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. 14025
    Prev
    Next
    INDEX
    Explanations

    phrases related to specialization or expertise in a field

    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
    383
    +0.13
    0.6%
    575
    +0.09
    0.4%
    673
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    383
    +0.13
    0.02
    624
    +0.09
    0.02
    390
    +0.09
    0.01
    Negative Logits
     Carth
    -0.66
     inev
    -0.62
     Jed
    -0.61
     Guth
    -0.60
     scramble
    -0.60
     desperation
    -0.59
     Palmer
    -0.56
     Decay
    -0.56
     Compton
    -0.56
     Downing
    -0.56
    POSITIVE LOGITS
    ately
    0.85
    ently
    0.83
    aciously
    0.82
    osed
    0.76
    inical
    0.76
    lems
    0.74
    heastern
    0.74
    inant
    0.74
    atography
    0.73
    itates
    0.73
    Activations Density 0.082%

    No Known Activations