© 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. 10934
    Prev
    Next
    INDEX
    Explanations

    sentences related to software or technology updates

    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
    480
    +0.38
    2.0%
    447
    +0.36
    1.9%
    481
    +0.20
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    266
    +0.38
    0.08
    480
    +0.36
    0.03
    276
    +0.20
    0.08
    Negative Logits
    omever
    -0.69
    ÃĥÃĤ
    -0.67
     administr
    -0.64
    itored
    -0.63
    anamo
    -0.62
     advoc
    -0.60
    elling
    -0.59
    ãĥ¼ãĥĨ
    -0.58
    ocrates
    -0.58
    iety
    -0.57
    POSITIVE LOGITS
     anime
    0.61
     playable
    0.60
     Xperia
    0.60
     manga
    0.57
     ASUS
    0.57
     Xiaomi
    0.57
     multiplayer
    0.57
     teaser
    0.56
     TTC
    0.55
     OnePlus
    0.54
    Activations Density 0.634%

    No Known Activations