© 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. 20869
    Prev
    Next
    INDEX
    Explanations

    neighborhood-related incidents or descriptive terms related to community violence or disturbances

    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
    48
    +0.12
    0.5%
    38
    +0.11
    0.5%
    746
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    555
    +0.12
    0.02
    48
    +0.11
    0.02
    326
    +0.10
    0.02
    Negative Logits
     caution
    -0.75
     Leap
    -0.70
    ISO
    -0.70
     disembark
    -0.65
     bearer
    -0.62
     Polar
    -0.62
     cipher
    -0.62
     printing
    -0.62
     faint
    -0.61
     overboard
    -0.61
    POSITIVE LOGITS
    bors
    1.30
    bour
    1.20
    eties
    1.17
    ths
    1.15
    elong
    1.08
    etimes
    1.08
    ª
    1.06
    etime
    1.05
    uries
    1.04
    «
    1.04
    Activations Density 0.039%

    No Known Activations