© 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 49152
    9. 8-RES_FS49152-JB
    11. 26845
    Prev
    Next
    INDEX
    Explanations

    numbers and measurements related to statistics and quantities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    260
    +0.11
    0.5%
    71
    +0.11
    0.5%
    599
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    260
    +0.11
    0.05
    570
    +0.11
    0.02
    344
    +0.10
    0.05
    Negative Logits
    inventoryQuantity
    -0.72
    paren
    -0.62
     Dying
    -0.61
    alde
    -0.60
    rog
    -0.60
    IELD
    -0.59
     Ragnarok
    -0.58
    Bey
    -0.58
    Secondly
    -0.58
    too
    -0.56
    POSITIVE LOGITS
     apiece
    1.17
     respectively
    1.14
     depending
    0.95
     including
    0.91
     each
    0.85
     annually
    0.84
     spanning
    0.84
     ranging
    0.83
     consecut
    0.79
     totaling
    0.78
    Activations Density 0.225%

    No Known Activations