© 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. 12473
    Prev
    Next
    INDEX
    Explanations

    terms related to aspects of power and influence in a variety of contexts

    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
    532
    +0.10
    0.5%
    762
    +0.10
    0.5%
    465
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    225
    +0.10
    0.06
    575
    +0.10
    0.06
    762
    +0.10
    0.07
    Negative Logits
    whose
    -0.63
    ossal
    -0.61
    ophon
    -0.60
    oln
    -0.60
     occasions
    -0.59
    ored
    -0.59
    missible
    -0.58
    rived
    -0.56
    arah
    -0.56
    catentry
    -0.56
    POSITIVE LOGITS
     wisely
    1.08
     accordingly
    0.88
     peacefully
    0.88
     selves
    0.81
     responsibly
    0.77
     carefully
    0.76
     elsewhere
    0.75
     cautiously
    0.74
     indoors
    0.74
     faithfully
    0.73
    Activations Density 0.311%

    No Known Activations