© 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. 20672
    Prev
    Next
    INDEX
    Explanations

    URL links and terms related to accessing files, information, or services online

    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
    497
    +0.15
    0.7%
    218
    +0.13
    0.6%
    323
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    323
    +0.15
    0.02
    271
    +0.13
    0.01
    1
    +0.12
    0.01
    Negative Logits
    zbollah
    -0.67
    deal
    -0.61
    fare
    -0.61
    plom
    -0.59
     quart
    -0.59
    warming
    -0.58
    consider
    -0.58
    friend
    -0.57
    stad
    -0.57
     circling
    -0.57
    POSITIVE LOGITS
    ibility
    1.13
    ibly
    1.11
    ible
    0.88
    ioned
    0.88
    IBLE
    0.85
    folios
    0.80
    ibles
    0.80
    ured
    0.77
    ories
    0.76
    itures
    0.75
    Activations Density 0.031%

    No Known Activations