© 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. 659
    Prev
    Next
    INDEX
    Explanations

    statistics and data related to labor trends and workforce demographics

    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
    165
    +0.10
    0.5%
    596
    +0.10
    0.4%
    576
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    488
    +0.10
    0.13
    576
    +0.10
    0.06
    607
    +0.09
    0.10
    Negative Logits
     tink
    -0.71
     propri
    -0.70
     oath
    -0.70
     coward
    -0.70
     authent
    -0.68
     blasp
    -0.68
     philosophies
    -0.68
     guilt
    -0.67
     traitor
    -0.67
     incarn
    -0.67
    POSITIVE LOGITS
     median
    1.35
    Ranked
    1.24
     Percentage
    1.23
    Average
    1.22
     percentage
    1.14
     average
    1.13
    Percent
    1.13
    adjusted
    1.12
     Median
    1.12
     Average
    1.11
    Activations Density 8.300%

    No Known Activations