© 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. 14263
    Prev
    Next
    INDEX
    Explanations

    conjunctions such as "and" and numerical expressions like "five up and five down"

    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
    519
    +0.10
    0.4%
    746
    +0.10
    0.4%
    374
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    746
    +0.10
    0.02
    427
    +0.10
    0.02
    484
    +0.10
    0.02
    Negative Logits
    weekly
    -0.75
    Week
    -0.72
     Helpful
    -0.70
    uel
    -0.70
    etheless
    -0.69
     Weekly
    -0.68
    æѦ
    -0.68
    ZA
    -0.67
    nor
    -0.67
    é¾
    -0.65
    POSITIVE LOGITS
     downs
    0.85
     scratch
    0.75
     down
    0.68
     sweaty
    0.64
    rew
    0.64
    aunder
    0.63
     forth
    0.62
     gown
    0.62
    rogens
    0.61
     saddle
    0.61
    Activations Density 0.075%

    No Known Activations