© 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. 5266
    Prev
    Next
    INDEX
    Explanations

    personal pronouns

    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
    481
    +0.25
    1.2%
    579
    +0.12
    0.6%
    115
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.25
    0.06
    579
    +0.12
    0.05
    232
    +0.10
    0.05
    Negative Logits
    -+-+
    -0.77
     conclud
    -0.74
     rall
    -0.72
     skelet
    -0.70
    STDOUT
    -0.70
    20439
    -0.66
     fec
    -0.63
     inflamm
    -0.63
    ICES
    -0.63
     Deadpool
    -0.61
    POSITIVE LOGITS
    ji
    1.24
    jin
    1.09
    ya
    1.07
    cki
    1.07
    emi
    1.04
    hu
    1.01
    wu
    1.00
    ui
    1.00
    uru
    0.99
    hai
    0.99
    Activations Density 0.182%

    No Known Activations