© 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. 22758
    Prev
    Next
    INDEX
    Explanations

    phrases related to social justice and marginalized communities

    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
    640
    +0.33
    1.6%
    314
    +0.12
    0.6%
    532
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.33
    0.15
    94
    +0.12
    0.09
    143
    +0.09
    0.07
    Negative Logits
     endeav
    -0.78
     endeavour
    -0.76
     opting
    -0.72
    hement
    -0.70
     diving
    -0.69
     dissu
    -0.67
    cffff
    -0.66
     undermin
    -0.66
     hashing
    -0.65
     engagements
    -0.65
    POSITIVE LOGITS
    Posted
    1.31
    posted
    1.27
    ³³³³³³³³
    1.25
    ³³³³³³³³³³³³³³³³
    1.15
    ³³³
    1.13
    ³³³³
    1.11
    ³³
    0.90
    rawdownloadcloneembedreportprint
    0.85
    RAW
    0.82
    Trivia
    0.80
    Activations Density 0.393%

    No Known Activations