© 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. 10382
    Prev
    Next
    INDEX
    Explanations

    references to specific things/events or detailed descriptions related to incidents or activities

    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
    314
    +0.13
    0.6%
    641
    +0.10
    0.5%
    515
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.09
    641
    +0.10
    0.06
    651
    +0.10
    0.06
    Negative Logits
    creen
    -0.82
    cale
    -0.70
    rica
    -0.66
    hips
    -0.65
    uckland
    -0.65
    zona
    -0.65
     straw
    -0.65
    hirt
    -0.65
    nesday
    -0.63
     Lank
    -0.63
    POSITIVE LOGITS
    ally
    0.95
    less
    0.93
    ually
    0.92
    ality
    0.91
    wise
    0.90
     wise
    0.86
    ual
    0.84
    lessly
    0.82
    Maker
    0.81
    eering
    0.78
    Activations Density 0.592%

    No Known Activations