© 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. 18650
    Prev
    Next
    INDEX
    Explanations

    mentions of the word "soft" with varying intensities

    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
    710
    +0.13
    0.6%
    745
    +0.11
    0.5%
    575
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    710
    +0.13
    0.02
    65
    +0.11
    0.02
    622
    +0.11
    0.01
    Negative Logits
    agher
    -0.73
    reon
    -0.72
    ulhu
    -0.70
    OUGH
    -0.70
    naires
    -0.67
     Ancients
    -0.64
     Pax
    -0.63
    IVERS
    -0.62
    DEV
    -0.61
     Forever
    -0.61
    POSITIVE LOGITS
    ening
    1.24
    ener
    1.21
    ball
    1.17
    hearted
    1.05
    eners
    1.04
    cover
    1.00
     palate
    0.97
    ened
    0.95
    balls
    0.95
    est
    0.91
    Activations Density 0.013%

    No Known Activations