© 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. 22589
    Prev
    Next
    INDEX
    Explanations

    phrases related to consequences or conditions that begin with "for."

    oai_token-act-pair · gpt-4o-miniTriggered 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
    620
    +0.10
    0.4%
    248
    +0.09
    0.4%
    414
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.10
    0.06
    641
    +0.09
    0.05
    572
    +0.09
    0.04
    Negative Logits
    orbit
    -0.69
    operated
    -0.62
    MSN
    -0.62
    hess
    -0.61
    #$
    -0.57
    Math
    -0.56
    buster
    -0.56
    bern
    -0.56
    jet
    -0.56
    heimer
    -0.56
    POSITIVE LOGITS
     purposes
    1.09
    gotten
    1.09
     sake
    1.03
     ourselves
    1.01
    bidden
    0.97
     awhile
    0.97
     oneself
    0.97
     example
    0.94
     eternity
    0.94
     themselves
    0.93
    Activations Density 0.179%

    No Known Activations