© 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. 18437
    Prev
    Next
    INDEX
    Explanations

    words related to sharing personal experiences and activities, particularly ones involving technology and communication

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    87
    +0.15
    0.7%
    224
    +0.11
    0.5%
    503
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    225
    +0.15
    0.07
    575
    +0.11
    0.08
    665
    +0.09
    0.07
    Negative Logits
    \":
    -0.58
    à©
    -0.57
    BuyableInstoreAndOnline
    -0.51
     hindsight
    -0.51
     à¨
    -0.51
     partName
    -0.50
     Analog
    -0.49
    inventoryQuantity
    -0.48
     Moz
    -0.48
    nces
    -0.48
    POSITIVE LOGITS
     safely
    0.87
    ().
    0.81
     purposes
    0.72
     efficiently
    0.70
    ãĢĤ
    0.69
    .
    0.69
     desired
    0.68
     quicker
    0.67
    $.
    0.66
    .<
    0.65
    Activations Density 0.818%

    No Known Activations