© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3144
    Prev
    Next
    INDEX
    Explanations

    descriptions of specific events, incidents, or occurrences

    oai_token-act-pair · gpt-3.5-turbo
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    481
    +0.22
    1.1%
    393
    +0.16
    0.8%
    640
    +0.15
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    385
    +0.22
    0.10
    362
    +0.16
    0.09
    608
    +0.15
    0.09
    Negative Logits
    ÏĢ
    -0.81
    \<
    -0.79
    Ïī
    -0.75
    .</
    -0.74
    ptr
    -0.74
    ceive
    -0.73
    hops
    -0.71
    ����
    -0.71
    !.
    -0.70
    åĤ
    -0.69
    POSITIVE LOGITS
    resa
    1.40
    odore
    1.36
    oret
    1.25
     latest
    1.18
     latter
    1.04
    ories
    1.01
     biggest
    0.98
     announcement
    0.97
     irony
    0.97
     idea
    0.95
    Activations Density 0.337%

    No Known Activations