© 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. 21138
    Prev
    Next
    INDEX
    Explanations

    phrases expressing an exaggerated or intense feeling, often related to negativity or criticism

    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
    373
    +0.16
    0.7%
    139
    +0.10
    0.5%
    565
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.16
    0.02
    132
    +0.10
    0.01
    565
    +0.10
    0.01
    Negative Logits
    cius
    -0.80
    instein
    -0.76
    yout
    -0.73
    yip
    -0.71
    igs
    -0.71
    aea
    -0.69
    76561
    -0.69
    yon
    -0.67
    etsk
    -0.66
     Mole
    -0.65
    POSITIVE LOGITS
     evident
    0.73
     eager
    0.73
     painfully
    0.71
     abundantly
    0.71
     prevalent
    0.70
    Much
    0.64
     Nguyen
    0.64
    BuyableInstoreAndOnline
    0.63
     often
    0.63
    entimes
    0.62
    Activations Density 0.024%

    No Known Activations