© 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. 10228
    Prev
    Next
    INDEX
    Explanations

    mentions of tasks or actions that need to be carried out

    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
    481
    +0.26
    1.2%
    765
    +0.12
    0.6%
    374
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.26
    0.04
    78
    +0.12
    0.03
    762
    +0.10
    0.04
    Negative Logits
    CHAT
    -0.76
    pered
    -0.68
    rin
    -0.67
    uffed
    -0.65
    ItemTracker
    -0.63
    uses
    -0.59
    irl
    -0.59
    usra
    -0.58
    ricted
    -0.58
     opin
    -0.57
    POSITIVE LOGITS
     anytime
    0.92
    uate
    0.91
     ASAP
    0.79
     regardless
    0.77
     whenever
    0.76
     before
    0.75
     regarding
    0.73
     sooner
    0.73
     BEFORE
    0.73
     anyways
    0.72
    Activations Density 0.124%

    No Known Activations