© 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. 8517
    Prev
    Next
    INDEX
    Explanations

    names of specific places or organizations, especially related to legal or political matters

    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
    314
    +0.13
    0.6%
    758
    +0.10
    0.5%
    367
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    376
    +0.13
    0.02
    201
    +0.10
    0.02
    367
    +0.10
    0.02
    Negative Logits
    ãĥ¼ãĥĨãĤ£
    -0.81
    olicy
    -0.78
    æĪ¦
    -0.74
    ãģį
    -0.72
     sterling
    -0.71
     Chaser
    -0.69
     Tycoon
    -0.69
    soDeliveryDate
    -0.68
     Mechdragon
    -0.66
    ãģ¯
    -0.65
    POSITIVE LOGITS
    manac
    1.27
    gorithm
    1.22
    ibaba
    1.22
    gebra
    1.20
    aska
    1.18
    cohol
    1.17
    phabet
    1.17
    gorith
    1.17
    chemy
    1.16
    umni
    1.12
    Activations Density 0.024%

    No Known Activations