© 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. 17967
    Prev
    Next
    INDEX
    Explanations

    mentions of talking or communicating with someone in a negative context

    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
    97
    +0.13
    0.6%
    640
    +0.09
    0.4%
    235
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    97
    +0.13
    0.03
    380
    +0.09
    0.02
    695
    +0.09
    0.02
    Negative Logits
    pite
    -0.68
    gradation
    -0.62
    gross
    -0.61
     ILCS
    -0.61
    Ranked
    -0.60
    ebted
    -0.60
    fighter
    -0.60
    wcsstore
    -0.58
    coming
    -0.58
    nav
    -0.58
    POSITIVE LOGITS
     him
    0.92
     reporters
    0.77
     strangers
    0.75
     representatives
    0.71
     them
    0.71
     passers
    0.70
    è¦ļéĨĴ
    0.69
     us
    0.67
     assistants
    0.65
     someone
    0.65
    Activations Density 0.059%

    No Known Activations