© 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. 292
    Prev
    Next
    INDEX
    Explanations

    phrases indicating significance or importance of something

    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
    742
    +0.12
    0.5%
    679
    +0.10
    0.5%
    447
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    540
    +0.12
    0.05
    737
    +0.10
    0.04
    324
    +0.09
    0.04
    Negative Logits
     vomit
    -0.72
     crabs
    -0.69
     worms
    -0.65
     butterflies
    -0.64
     Mush
    -0.64
     cages
    -0.63
     spiders
    -0.62
     roomm
    -0.61
     crab
    -0.61
    speak
    -0.60
    POSITIVE LOGITS
     helping
    0.86
     motivating
    0.79
    âĹ¼
    0.79
     determining
    0.78
     influencing
    0.76
     shaping
    0.74
     facilitating
    0.74
     contributing
    0.73
     towards
    0.72
     toward
    0.71
    Activations Density 0.188%

    No Known Activations