© 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. 751
    Prev
    Next
    INDEX
    Explanations

    references to Air Force-related terms and activities

    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
    459
    +0.13
    0.6%
    512
    +0.10
    0.5%
    179
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.13
    0.20
    326
    +0.10
    0.15
    289
    +0.09
    0.02
    Negative Logits
    umen
    -0.81
     Cumber
    -0.77
    inois
    -0.72
    ublic
    -0.71
    FINE
    -0.71
     Staples
    -0.67
    usat
    -0.67
    hesis
    -0.67
    utics
    -0.66
    illard
    -0.65
    POSITIVE LOGITS
    plane
    1.27
    planes
    1.16
    walk
    1.00
    hawk
    0.99
     Flight
    0.94
    flight
    0.94
     refuel
    0.94
     hangar
    0.93
    walker
    0.92
    crew
    0.91
    Activations Density 4.534%

    No Known Activations