© 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. 10137
    Prev
    Next
    INDEX
    Explanations

    emails and specific days of the week

    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
    280
    +0.13
    0.6%
    713
    +0.10
    0.5%
    225
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    503
    +0.13
    0.04
    212
    +0.10
    0.05
    597
    +0.10
    0.05
    Negative Logits
     consequential
    -0.73
     secondary
    -0.66
     generic
    -0.65
     Secondary
    -0.64
     attribution
    -0.64
     sectarian
    -0.62
     Gothic
    -0.62
     inclusive
    -0.61
     preparation
    -0.61
     remaining
    -0.61
    POSITIVE LOGITS
    wrote
    1.40
    igrated
    1.16
    ounced
    1.15
    ifies
    1.11
    ulates
    1.08
    izes
    1.08
    heartedly
    1.07
    overed
    1.04
    elled
    1.03
    did
    1.03
    Activations Density 0.539%

    No Known Activations