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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 1536
    9. 8-RES_FS1536-JB
    11. 71
    Prev
    Next
    INDEX
    Explanations

    words related to international conflicts and political tensions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    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
    649
    +0.13
    0.6%
    516
    +0.10
    0.5%
    436
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.11
    289
    +0.10
    0.03
    346
    +0.10
    0.07
    Negative Logits
    ttes
    -0.81
    utic
    -0.78
    mson
    -0.77
    ĸļ
    -0.70
    parts
    -0.69
     Gators
    -0.68
    teenth
    -0.67
    isters
    -0.66
     NCT
    -0.64
     Bleach
    -0.64
    POSITIVE LOGITS
    anyahu
    1.13
     Aviv
    1.04
    ×Ļ×
    0.95
     ×
    0.88
     Netanyahu
    0.86
    itz
    0.78
     Judah
    0.77
    ×
    0.76
    aretz
    0.74
     Lobby
    0.72
    Activations Density 2.002%

    No Known Activations