© 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. 16163
    Prev
    Next
    INDEX
    Explanations

    names of specific entities such as teams, companies, symbols, fonts, laws, religions, political institutions, races, artists, albums, and specific concepts like temperature and faults

    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
    447
    +0.15
    0.7%
    481
    +0.13
    0.6%
    19
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.15
    0.07
    19
    +0.13
    0.05
    40
    +0.11
    0.05
    Negative Logits
     Supplemental
    -0.61
    hang
    -0.60
    bah
    -0.59
    ensive
    -0.58
     Bere
    -0.58
    athered
    -0.58
    ================================
    -0.57
     largeDownload
    -0.57
    ounded
    -0.56
    nyder
    -0.55
    POSITIVE LOGITS
     imaginable
    1.00
     or
    0.76
     descriptor
    0.74
    Id
    0.74
     whatsoever
    0.70
     combinations
    0.69
     bidder
    0.69
     identifier
    0.67
     affiliation
    0.67
     besides
    0.67
    Activations Density 0.331%

    No Known Activations