© 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. 723
    Prev
    Next
    INDEX
    Explanations

    programming-related terms and commands

    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
    700
    +0.12
    0.6%
    255
    +0.12
    0.5%
    345
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    345
    +0.12
    0.05
    326
    +0.12
    0.04
    255
    +0.11
    0.04
    Negative Logits
    laus
    -0.80
    ayn
    -0.71
    guyen
    -0.70
    akening
    -0.68
     lawy
    -0.67
    doms
    -0.65
    undy
    -0.63
    arag
    -0.63
     Aven
    -0.63
    ights
    -0.61
    POSITIVE LOGITS
    ()
    1.09
    ():
    0.98
    able
    0.93
    ().
    0.92
    (),
    0.89
    ables
    0.89
    Transaction
    0.87
    Async
    0.85
    ();
    0.84
    ername
    0.83
    Activations Density 0.211%

    No Known Activations