© 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. 765
    Prev
    Next
    INDEX
    Explanations

    instructions or steps outlined in a document

    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
    665
    +0.14
    0.7%
    43
    +0.11
    0.5%
    468
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    665
    +0.14
    0.27
    70
    +0.11
    0.21
    761
    +0.11
    0.17
    Negative Logits
    pmwiki
    -0.64
     wont
    -0.63
     indicating
    -0.61
     marked
    -0.59
     perme
    -0.58
     suggesting
    -0.58
    ItemTracker
    -0.57
     favoring
    -0.55
    blogspot
    -0.55
    cean
    -0.54
    POSITIVE LOGITS
     oneself
    1.21
     them
    1.07
    uate
    1.05
     anything
    1.01
    ulate
    0.99
    urate
    0.95
     him
    0.92
    igate
    0.91
    enance
    0.90
     ourselves
    0.89
    Activations Density 5.785%

    No Known Activations