© 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 24576
    9. 8-RES_FS24576-JB
    11. 18894
    Prev
    Next
    INDEX
    Explanations

    terms related to the Mandelbrot set, including mentions of its specific parts, properties, and explanations

    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
    7
    +0.11
    0.5%
    376
    +0.11
    0.5%
    396
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    376
    +0.11
    0.01
    232
    +0.11
    0.01
    23
    +0.10
    0.01
    Negative Logits
    ãģį
    -0.79
     Typhoon
    -0.78
     Spit
    -0.71
    ï¸
    -0.69
    spection
    -0.68
     sights
    -0.66
     Depth
    -0.63
     FT
    -0.62
    ··
    -0.62
    Magikarp
    -0.62
    POSITIVE LOGITS
    olin
    1.25
    arin
    1.10
    eer
    1.09
    erer
    1.08
    ela
    1.05
    rag
    1.04
    rake
    1.03
    arine
    1.01
    atory
    0.99
    ered
    0.97
    Activations Density 0.027%

    No Known Activations