© 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. 5207
    Prev
    Next
    INDEX
    Explanations

    one-word actions with a strong positive or accomplished connotation

    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
    509
    +0.10
    0.4%
    38
    +0.10
    0.4%
    477
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    70
    +0.10
    0.06
    195
    +0.10
    0.02
    427
    +0.08
    0.04
    Negative Logits
    issance
    -0.68
     COURT
    -0.67
    iscons
    -0.64
     destro
    -0.62
    ammy
    -0.61
     msec
    -0.58
    Cub
    -0.58
     encount
    -0.58
    hin
    -0.57
     agre
    -0.56
    POSITIVE LOGITS
    ings
    1.08
    ingly
    1.07
    eth
    0.91
    fully
    0.88
    adelphia
    0.84
     themselves
    0.83
    able
    0.81
    ably
    0.80
    ned
    0.79
    ments
    0.77
    Activations Density 0.434%

    No Known Activations