© 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 768
    9. 8-RES_FS768-JB
    11. 661
    Prev
    Next
    INDEX
    Explanations

    direct quotations and reported speech

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs768-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    768
    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
    679
    +0.12
    0.5%
    514
    +0.09
    0.4%
    512
    +0.08
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.12
    0.11
    274
    +0.09
    0.12
    468
    +0.08
    0.11
    Negative Logits
    à¦
    -0.84
    Pont
    -0.82
    pee
    -0.71
    Holy
    -0.69
    irrel
    -0.68
    ãĥİ
    -0.66
    ãĥ¯
    -0.66
    kay
    -0.66
    gone
    -0.66
    pe
    -0.65
    POSITIVE LOGITS
     bluntly
    0.89
     afterward
    0.85
     sarcast
    0.84
     goodbye
    0.84
     anecd
    0.81
     afterwards
    0.80
     yesterday
    0.73
    :"
    0.72
     Wednesday
    0.70
     emphatically
    0.70
    Activations Density 2.222%

    No Known Activations