© 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. 17145
    Prev
    Next
    INDEX
    Explanations

    words related to lists or items, often preceded by phrases like "to name a few"

    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
    697
    +0.12
    0.5%
    155
    +0.10
    0.5%
    556
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    155
    +0.12
    0.02
    172
    +0.10
    0.02
    568
    +0.09
    0.02
    Negative Logits
    jon
    -0.79
    iry
    -0.72
    ysc
    -0.66
    nel
    -0.63
    hiro
    -0.63
     REPORT
    -0.62
    rus
    -0.60
    imum
    -0.59
    iol
    -0.59
    irt
    -0.58
    POSITIVE LOGITS
     favorites
    0.93
     names
    0.89
     favourites
    0.86
     examples
    0.83
     acron
    0.81
     favorite
    0.76
     favourite
    0.74
     specifics
    0.74
     adject
    0.73
     suffice
    0.71
    Activations Density 0.110%

    No Known Activations