© 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
    2. Joseph Bloom · Feature Splitting for GPT2-Small
    3. GPT2-Small
    4. Residual Feature Splitting 24576
    5. 8-RES_FS24576-JB
    6. 23872
    Prev
    Next
    INDEX
    Explanations

    references to chapters and sections in written works, particularly in a literary or academic context

    oai_token-act-pair · gpt-4o-miniTriggered 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
    420
    +0.10
    0.5%
    526
    +0.10
    0.5%
    113
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    326
    +0.10
    0.02
    288
    +0.10
    0.02
    232
    +0.10
    0.02
    Negative Logits
     intensity
    -0.61
    iquid
    -0.57
    ftime
    -0.57
     Voltage
    -0.56
     DEFENSE
    -0.56
    cular
    -0.55
     snap
    -0.55
    âķIJ
    -0.54
    liction
    -0.54
     offic
    -0.54
    POSITIVE LOGITS
    arios
    0.76
    stones
    0.74
    eless
    0.73
    flies
    0.73
    sites
    0.71
    stakes
    0.70
    glers
    0.69
    ENG
    0.69
    book
    0.69
    del
    0.68
    Activations Density 0.048%

    No Known Activations