© 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 12288
    9. 8-RES_FS12288-JB
    11. 10380
    Prev
    Next
    INDEX
    Explanations

    mentions of appreciation or criticism

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    223
    +0.11
    0.5%
    13
    +0.11
    0.5%
    764
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    13
    +0.11
    0.05
    567
    +0.11
    0.04
    223
    +0.10
    0.04
    Negative Logits
    iba
    -0.75
    CHA
    -0.65
    Ct
    -0.64
    ADE
    -0.62
    yrinth
    -0.60
    ODY
    -0.60
    aten
    -0.59
    NING
    -0.59
    edIn
    -0.59
    ourge
    -0.58
    POSITIVE LOGITS
     impression
    0.94
     thumbs
    0.91
    icum
    0.81
     advice
    0.77
     reins
    0.77
     nod
    0.76
     impetus
    0.75
     insight
    0.74
     keynote
    0.73
     cred
    0.73
    Activations Density 0.317%

    No Known Activations