© 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. 18870
    Prev
    Next
    INDEX
    Explanations

    proper nouns

    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
    87
    +0.22
    1.1%
    481
    +0.10
    0.5%
    146
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    146
    +0.22
    0.06
    698
    +0.10
    0.05
    691
    +0.09
    0.05
    Negative Logits
    rote
    -0.71
    Berry
    -0.57
     Mehran
    -0.56
    rotein
    -0.55
    urized
    -0.55
    rams
    -0.53
    geon
    -0.52
     Peel
    -0.52
     Audi
    -0.52
    actionDate
    -0.52
    POSITIVE LOGITS
     being
    1.26
     needing
    1.18
     becoming
    1.12
     having
    1.06
     disappearing
    1.06
     wanting
    1.05
     interfering
    1.02
    being
    1.02
     losing
    1.00
     failing
    1.00
    Activations Density 0.662%

    No Known Activations