© 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. 22459
    Prev
    Next
    INDEX
    Explanations

    connective phrases and conjunctions

    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
    447
    +0.14
    0.6%
    714
    +0.12
    0.5%
    420
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    471
    +0.14
    0.05
    546
    +0.12
    0.05
    720
    +0.11
    0.05
    Negative Logits
    rir
    -0.74
    UAL
    -0.72
    dayName
    -0.71
    appropriate
    -0.70
    ital
    -0.65
    bal
    -0.65
    imeter
    -0.65
    Specifically
    -0.63
    ept
    -0.63
    Favorite
    -0.62
    POSITIVE LOGITS
     replaced
    1.34
     reborn
    1.02
     vanished
    0.99
     disappeared
    0.99
     leaving
    0.93
     reapp
    0.90
     rebuilt
    0.89
     Leaves
    0.88
     exile
    0.87
     forgotten
    0.86
    Activations Density 0.534%

    No Known Activations