© 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. 5545
    Prev
    Next
    INDEX
    Explanations

    information related to social media posts and political events or statements

    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
    87
    +0.15
    0.7%
    475
    +0.13
    0.6%
    679
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    470
    +0.15
    0.05
    195
    +0.13
    0.03
    679
    +0.12
    0.08
    Negative Logits
    .","
    -1.01
     guiActiveUnfocused
    -0.81
    aches
    -0.75
    \">
    -0.75
    aceae
    -0.73
    sg
    -0.73
    "},"
    -0.70
     biod
    -0.66
     apprentices
    -0.66
    olesc
    -0.64
    POSITIVE LOGITS
     retweet
    1.02
     irony
    0.96
     rebutt
    0.93
     debunked
    0.92
     commenter
    0.91
     Huckabee
    0.90
     Breitbart
    0.89
     Coulter
    0.89
     hypocrisy
    0.89
     redacted
    0.88
    Activations Density 1.024%

    No Known Activations