© 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. 19862
    Prev
    Next
    INDEX
    Explanations

    proper nouns related to various subjects such as technology, medicine, and religious texts

    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
    481
    +0.52
    2.8%
    447
    +0.26
    1.4%
    480
    +0.14
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    659
    +0.52
    0.07
    326
    +0.26
    0.06
    232
    +0.14
    0.06
    Negative Logits
     conduc
    -0.65
     Citiz
    -0.62
     advoc
    -0.60
     accomp
    -0.59
     trave
    -0.58
     internationally
    -0.57
     Ezek
    -0.56
    ij士
    -0.55
     encount
    -0.54
     agre
    -0.54
    POSITIVE LOGITS
    imum
    0.71
    jit
    0.69
    adium
    0.69
    tera
    0.69
    oba
    0.67
    dylib
    0.67
    jac
    0.67
    ipop
    0.67
    weed
    0.66
    heim
    0.65
    Activations Density 0.394%

    No Known Activations