© 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 EvalsExportsAPI 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. 18031
    Prev
    Next
    INDEX
    Explanations

    names of political figures or influential individuals

    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
    314
    +0.14
    0.6%
    651
    +0.11
    0.5%
    710
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.14
    0.02
    289
    +0.11
    0.01
    442
    +0.10
    0.01
    Negative Logits
    ©¶æ
    -0.74
     fracturing
    -0.62
     Vaugh
    -0.57
     curv
    -0.56
    indal
    -0.56
     wills
    -0.56
     offic
    -0.55
    shire
    -0.54
    Blade
    -0.54
     unstoppable
    -0.54
    POSITIVE LOGITS
    terson
    1.17
    rice
    1.13
    rix
    1.11
    ting
    1.03
    herer
    1.00
    ches
    0.99
    chet
    0.99
    alon
    0.98
    rons
    0.96
    hemat
    0.96
    Activations Density 0.040%

    No Known Activations