© 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. 7912
    Prev
    Next
    INDEX
    Explanations

    last names ending in "ent" or "ent" itself

    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
    147
    +0.13
    0.6%
    284
    +0.12
    0.6%
    311
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.13
    0.02
    404
    +0.12
    0.02
    147
    +0.12
    0.02
    Negative Logits
    BILITIES
    -0.78
    STON
    -0.75
    issance
    -0.68
    STER
    -0.65
    iets
    -0.64
    BILITY
    -0.63
    side
    -0.63
     Heist
    -0.63
     Scythe
    -0.62
    CHR
    -0.61
    POSITIVE LOGITS
    ucky
    1.13
    ral
    1.12
    ropy
    1.12
    inel
    1.00
    rification
    0.96
    ertain
    0.95
    acles
    0.95
    acion
    0.94
    reprene
    0.94
    imental
    0.93
    Activations Density 0.027%

    No Known Activations