© 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 24576
    9. 8-RES_FS24576-JB
    11. 24223
    Prev
    Next
    INDEX
    Explanations

    instances of the word "invitation" and its derivatives, such as "invocations."

    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
    526
    +0.11
    0.5%
    162
    +0.10
    0.5%
    274
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    162
    +0.11
    0.01
    369
    +0.10
    0.01
    232
    +0.10
    0.01
    Negative Logits
     cleaner
    -0.73
    Ħ¢
    -0.72
     bare
    -0.70
     horizontally
    -0.68
     fab
    -0.68
    Reloaded
    -0.63
    Ĥ¬
    -0.61
     showing
    -0.61
     Whitman
    -0.61
    beard
    -0.60
    POSITIVE LOGITS
    igor
    1.42
    itational
    1.41
    itations
    1.39
    asion
    1.38
    incible
    1.31
    itation
    1.30
    ocation
    1.30
    asions
    1.29
    ocations
    1.22
    idious
    1.20
    Activations Density 0.014%

    No Known Activations