© 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. 21084
    Prev
    Next
    INDEX
    Explanations

    descriptions related to the sense of taste

    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
    128
    +0.10
    0.5%
    195
    +0.10
    0.5%
    211
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    211
    +0.10
    0.01
    129
    +0.10
    0.01
    255
    +0.10
    0.01
    Negative Logits
     Freeman
    -0.69
    vernment
    -0.63
     Lak
    -0.63
    doms
    -0.63
     Rhod
    -0.61
    kson
    -0.61
     Occupations
    -0.61
    channelAvailability
    -0.60
     Spur
    -0.60
     sled
    -0.60
    POSITIVE LOGITS
     buds
    1.39
     tastes
    0.95
     palate
    0.94
     tasting
    0.93
     taste
    0.92
    terness
    0.85
     juices
    0.83
    hend
    0.81
     pleasures
    0.81
     smelling
    0.80
    Activations Density 0.023%

    No Known Activations