© 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. 5170
    Prev
    Next
    INDEX
    Explanations

    words related to physical exercise techniques and instructions

    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
    +0.13
    0.6%
    86
    +0.12
    0.6%
    49
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    86
    +0.13
    0.03
    762
    +0.12
    0.03
    354
    +0.10
    0.03
    Negative Logits
    76561
    -0.79
    REDACTED
    -0.69
    ULAR
    -0.68
    ASED
    -0.67
     ILCS
    -0.64
    BuyableInstoreAndOnline
    -0.64
    breeding
    -0.63
    edIn
    -0.61
     Sands
    -0.60
     Ago
    -0.59
    POSITIVE LOGITS
     cope
    1.28
     achieve
    1.26
     communicate
    1.18
     navigate
    1.17
     formulate
    1.11
     organize
    1.10
     improve
    1.09
     survive
    1.08
     recover
    1.07
     solve
    1.06
    Activations Density 0.100%

    No Known Activations