© 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. 2710
    Prev
    Next
    INDEX
    Explanations

    phrases related to controversial political statements or conspiracy theories

    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
    719
    +0.11
    0.5%
    314
    +0.10
    0.5%
    331
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    719
    +0.11
    0.06
    621
    +0.10
    0.05
    253
    +0.10
    0.05
    Negative Logits
     zoning
    -0.79
     territ
    -0.78
     plat
    -0.76
     hemor
    -0.75
     destro
    -0.72
     dust
    -0.70
     Marketable
    -0.70
     semblance
    -0.69
     charm
    -0.68
     Borough
    -0.67
    POSITIVE LOGITS
    ª
    1.23
    ¹
    1.21
    ¸
    1.07
    ı
    1.07
    «
    1.00
    ¦
    0.98
    ¾
    0.97
    taboola
    0.97
    ³
    0.97
    º
    0.97
    Activations Density 0.094%

    No Known Activations