© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 6758
    Prev
    Next
    INDEX
    Explanations

    specific names, titles, and political positions

    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
    +0.66
    4.4%
    87
    +0.26
    1.7%
    266
    +0.20
    1.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    480
    +0.66
    0.03
    266
    +0.26
    0.08
    232
    +0.20
    0.08
    Negative Logits
     destro
    -0.56
     disse
    -0.51
     anecd
    -0.48
    kson
    -0.46
     skelet
    -0.45
     slic
    -0.44
     pestic
    -0.44
     unnecess
    -0.44
    cms
    -0.44
    CONCLUS
    -0.43
    POSITIVE LOGITS
     Mans
    0.48
     ÂŃ
    0.47
     Donald
    0.43
     Tim
    0.43
     Min
    0.42
     Vernon
    0.40
     Ma
    0.40
     Ming
    0.39
     George
    0.39
     Maj
    0.39
    Activations Density 0.791%

    No Known Activations