© 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. 19875
    Prev
    Next
    INDEX
    Explanations

    the word "a" followed by adjectives

    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
    36
    +0.17
    0.8%
    393
    +0.15
    0.7%
    362
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.17
    0.06
    436
    +0.15
    0.03
    263
    +0.12
    0.03
    Negative Logits
    Edit
    -0.78
     nonetheless
    -0.73
    çīĪ
    -0.71
    Apparently
    -0.71
    Links
    -0.69
    unknown
    -0.68
     instead
    -0.67
    olicy
    -0.66
    edit
    -0.65
    almost
    -0.64
    POSITIVE LOGITS
     coincidence
    1.22
     anymore
    1.04
     pleasant
    0.91
     slightest
    0.91
     mere
    0.90
     nor
    0.90
     necessarily
    0.89
    verse
    0.88
     bother
    0.85
     viable
    0.84
    Activations Density 0.074%

    No Known Activations