© 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. 5300
    Prev
    Next
    INDEX
    Explanations

    Spanish accented letters such as "ñ" and specific terms related to Spanish-speaking countries

    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
    587
    +0.11
    0.5%
    7
    +0.11
    0.5%
    182
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    545
    +0.11
    0.01
    284
    +0.11
    0.01
    182
    +0.10
    0.01
    Negative Logits
     MISS
    -0.72
     Sapp
    -0.65
     Shutterstock
    -0.61
     fertility
    -0.61
     microsc
    -0.60
     SPACE
    -0.60
     abdom
    -0.60
    HEAD
    -0.60
    leigh
    -0.60
     magnification
    -0.60
    POSITIVE LOGITS
    ñ
    1.15
    ña
    1.10
    ez
    1.05
    rez
    0.91
    ón
    0.88
    ã
    0.84
    uel
    0.82
    untu
    0.82
    rette
    0.82
    illo
    0.81
    Activations Density 0.012%

    No Known Activations