© 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. 16156
    Prev
    Next
    INDEX
    Explanations

    words related to chemistry and laboratory procedures

    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
    339
    +0.15
    0.7%
    216
    +0.12
    0.5%
    19
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    339
    +0.15
    0.02
    659
    +0.12
    0.02
    155
    +0.10
    0.02
    Negative Logits
    Reviewer
    -0.73
    culosis
    -0.69
    Dub
    -0.67
    ilings
    -0.67
    accompanied
    -0.67
    BOX
    -0.65
    gee
    -0.65
    vernment
    -0.65
    perty
    -0.63
    女
    -0.63
    POSITIVE LOGITS
    ophe
    1.13
    orical
    1.04
    otle
    1.01
    ribution
    0.99
    rict
    0.99
    ocrat
    0.98
    ocene
    0.93
    opher
    0.91
    ricting
    0.90
    ribut
    0.89
    Activations Density 0.040%

    No Known Activations