© 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 12288
    9. 8-RES_FS12288-JB
    11. 10365
    Prev
    Next
    INDEX
    Explanations

    instances of word "of" followed by a significant term

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs12288-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    12,288
    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
    53
    +0.15
    0.7%
    393
    +0.11
    0.5%
    674
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    53
    +0.15
    0.05
    331
    +0.11
    0.03
    281
    +0.11
    0.02
    Negative Logits
    essors
    -0.72
    gyn
    -0.72
    uations
    -0.66
    ubs
    -0.66
    raints
    -0.65
    oother
    -0.62
    ilings
    -0.62
    irl
    -0.61
    annah
    -0.61
    atu
    -0.61
    POSITIVE LOGITS
     cloth
    0.83
     scenery
    0.83
     legislation
    0.81
     trivia
    0.79
     machinery
    0.78
     gum
    0.78
     wreckage
    0.77
     advice
    0.76
     luck
    0.75
     cake
    0.74
    Activations Density 0.105%

    No Known Activations