© 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. 1540
    Prev
    Next
    INDEX
    Explanations

    terms related to legislation, regulations, and opinions

    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
    393
    +0.11
    0.5%
    534
    +0.11
    0.5%
    101
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    534
    +0.11
    0.04
    572
    +0.11
    0.04
    101
    +0.11
    0.05
    Negative Logits
     stub
    -0.72
     blew
    -0.70
     behaved
    -0.70
     excuse
    -0.70
    Required
    -0.70
    SPONSORED
    -0.69
    Synopsis
    -0.68
     breeze
    -0.66
     fit
    -0.66
    matter
    -0.65
    POSITIVE LOGITS
     Greenpeace
    0.84
     advocacy
    0.82
     consultancy
    0.82
     Aval
    0.81
     Ceres
    0.79
     Advoc
    0.76
     Global
    0.76
     consulting
    0.76
     Concern
    0.73
     Consulting
    0.73
    Activations Density 0.147%

    No Known Activations