© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 5533
    Prev
    Next
    INDEX
    Explanations

    terms related to improvement or enhancement

    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
    665
    +0.12
    0.6%
    560
    +0.11
    0.5%
    45
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    455
    +0.12
    0.03
    45
    +0.11
    0.02
    560
    +0.09
    0.02
    Negative Logits
     Watching
    -0.67
     absence
    -0.66
     skipping
    -0.66
    Reviewer
    -0.66
    hell
    -0.64
     goodbye
    -0.63
     weeping
    -0.62
     waiting
    -0.62
     aborted
    -0.62
     legion
    -0.61
    POSITIVE LOGITS
     differentiate
    1.15
     stimulate
    1.12
     communicate
    1.09
     educate
    1.07
     regulate
    1.07
    aunder
    1.06
     appreciate
    1.04
     utilize
    1.04
     relate
    1.03
     satisfy
    1.03
    Activations Density 0.048%

    No Known Activations