© 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. 10208
    Prev
    Next
    INDEX
    Explanations

    references to people's positions and titles

    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
    123
    +0.11
    0.5%
    430
    +0.10
    0.5%
    182
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    463
    +0.11
    0.04
    586
    +0.10
    0.03
    703
    +0.10
    0.02
    Negative Logits
    nia
    -0.85
    icity
    -0.84
    ibly
    -0.83
    icals
    -0.81
    cific
    -0.78
    gas
    -0.76
    ague
    -0.75
    utic
    -0.75
    antics
    -0.75
    acci
    -0.73
    POSITIVE LOGITS
     CEO
    1.21
     chief
    1.01
     propriet
    0.96
    CEO
    0.95
     Chief
    0.94
     Publisher
    0.92
     founding
    0.90
     Executive
    0.90
     editor
    0.90
     treasurer
    0.88
    Activations Density 0.163%

    No Known Activations