© 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 12288
    9. 8-RES_FS12288-JB
    11. 318
    Prev
    Next
    INDEX
    Explanations

    paragraphs related to policy or military activities

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    640
    +0.26
    1.2%
    314
    +0.18
    0.8%
    442
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.26
    0.08
    615
    +0.18
    0.04
    160
    +0.10
    0.05
    Negative Logits
    hement
    -0.80
     Krypt
    -0.70
     XIII
    -0.68
     glare
    -0.65
     Slug
    -0.64
     Tag
    -0.64
     flats
    -0.62
     Lug
    -0.61
     Faction
    -0.61
     Versus
    -0.60
    POSITIVE LOGITS
    ³³³³³³³³³³³³³³³³
    1.10
    ³³³³³³³³
    1.07
    ³³³
    0.91
    shall
    0.90
    ³³³³
    0.87
    ó
    0.80
    ever
    0.80
    bush
    0.80
    eq
    0.79
    nature
    0.79
    Activations Density 0.253%

    No Known Activations