© 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. 2708
    Prev
    Next
    INDEX
    Explanations

    mentions of legal or technology-related terms and concepts

    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
    277
    +0.11
    0.5%
    288
    +0.10
    0.5%
    616
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    277
    +0.11
    0.03
    294
    +0.10
    0.02
    628
    +0.10
    0.02
    Negative Logits
     Clause
    -0.78
    士
    -0.77
     Vander
    -0.74
     Mercy
    -0.73
     Petty
    -0.73
     Electric
    -0.69
     Pearce
    -0.68
     Darrell
    -0.68
     Bauer
    -0.67
     Duchess
    -0.67
    POSITIVE LOGITS
     cafes
    1.02
     cafe
    0.93
     caf
    0.87
     café
    0.87
    omnia
    0.84
    osphere
    0.82
    izens
    0.82
     gateway
    0.81
    wal
    0.81
    worms
    0.80
    Activations Density 0.026%

    No Known Activations