© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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 12288
    9. 8-RES_FS12288-JB
    11. 1643
    Prev
    Next
    INDEX
    Explanations

    phrases related to overcoming obstacles or challenges

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    +0.12
    0.6%
    323
    +0.09
    0.4%
    649
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    587
    +0.12
    0.04
    323
    +0.09
    0.03
    237
    +0.09
    0.03
    Negative Logits
    isted
    -0.78
    Ring
    -0.76
    CTV
    -0.75
    UTH
    -0.73
    gemony
    -0.72
    abet
    -0.71
    etus
    -0.70
    IST
    -0.67
    Ĵ
    -0.65
    aires
    -0.64
    POSITIVE LOGITS
     waivers
    0.83
     hoops
    0.72
     puberty
    0.71
    inventoryQuantity
    0.70
     clenched
    0.68
     intermedi
    0.67
     thick
    0.65
     bushes
    0.64
     adolescence
    0.64
    )=(
    0.64
    Activations Density 0.070%

    No Known Activations