© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 24018
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns and expressions related to ownership or control

    oai_token-act-pair · gpt-4o-miniTriggered 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
    48
    +0.11
    0.5%
    407
    +0.11
    0.5%
    636
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    636
    +0.11
    0.05
    362
    +0.11
    0.05
    526
    +0.11
    0.01
    Negative Logits
    ItemImage
    -0.84
    ItemThumbnailImage
    -0.76
    Sold
    -0.70
    BuyableInstoreAndOnline
    -0.70
    DragonMagazine
    -0.69
    âĹ¼
    -0.68
    RT
    -0.66
    ITNESS
    -0.65
     hereafter
    -0.65
     Afric
    -0.65
    POSITIVE LOGITS
     sleeves
    1.01
     feathers
    1.00
     eyel
    0.99
     nose
    0.97
     fingers
    0.96
     toes
    0.96
     noses
    0.94
     sleeve
    0.94
     chops
    0.93
     socks
    0.92
    Activations Density 0.168%

    No Known Activations