© 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. 5876
    Prev
    Next
    INDEX
    Explanations

    instances where the phrase "it's been" is being used

    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
    55
    +0.14
    0.7%
    102
    +0.11
    0.5%
    566
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.14
    0.07
    102
    +0.11
    0.06
    566
    +0.10
    0.05
    Negative Logits
    lies
    -0.74
    ierrez
    -0.69
    autions
    -0.69
    lication
    -0.67
    idental
    -0.64
    rouse
    -0.63
    ampa
    -0.63
    lip
    -0.62
    PHOTOS
    -0.61
    Profile
    -0.61
    POSITIVE LOGITS
     awhile
    1.02
     likened
    0.89
     proven
    0.87
    cffffcc
    0.87
     taken
    0.84
     hijacked
    0.82
     replaced
    0.81
     raining
    0.80
     upgraded
    0.80
     forgotten
    0.79
    Activations Density 0.105%

    No Known Activations