© 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 EvalsExportsAPI 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. 94
    Prev
    Next
    INDEX
    Explanations

    phrases related to specific lengths of time

    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
    274
    +0.12
    0.6%
    657
    +0.12
    0.5%
    651
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    657
    +0.12
    0.07
    651
    +0.12
    0.06
    258
    +0.11
    0.05
    Negative Logits
    ople
    -0.80
    ibrary
    -0.71
    achus
    -0.64
     contag
    -0.64
    ergic
    -0.63
    urn
    -0.62
    icc
    -0.62
    ech
    -0.60
    grave
    -0.60
    enburg
    -0.60
    POSITIVE LOGITS
     ago
    1.53
     Ago
    1.28
     olds
    1.00
     probation
    0.86
     apiece
    0.82
    manship
    0.80
    long
    0.76
     hiatus
    0.76
     overdue
    0.74
    ago
    0.71
    Activations Density 0.086%

    No Known Activations