© 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. 5665
    Prev
    Next
    INDEX
    Explanations

    phrases relating to sensory perception, specifically focusing on the sense of taste and touch

    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
    481
    +0.14
    0.7%
    356
    +0.10
    0.5%
    43
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    632
    +0.14
    0.04
    356
    +0.10
    0.03
    43
    +0.09
    0.03
    Negative Logits
    SourceFile
    -0.81
    ULAR
    -0.79
    ãĤ¼ãĤ¦ãĤ¹
    -0.70
    pecially
    -0.69
    arah
    -0.65
    ige
    -0.64
    ENN
    -0.64
    PA
    -0.62
    aled
    -0.61
    ni
    -0.61
    POSITIVE LOGITS
     unlike
    1.06
     alas
    0.98
     beware
    0.93
     there
    0.92
     contrary
    0.87
     despite
    0.86
     according
    0.86
     if
    0.82
     whereas
    0.82
     none
    0.82
    Activations Density 0.108%

    No Known Activations