© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 2700
    Prev
    Next
    INDEX
    Explanations

    names of specific individuals mentioned in a news article

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    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
    447
    +0.20
    0.9%
    481
    +0.17
    0.8%
    121
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    430
    +0.20
    0.08
    111
    +0.17
    0.06
    154
    +0.09
    0.06
    Negative Logits
    ãĤ¡
    -0.73
    Americ
    -0.70
    CRE
    -0.69
    soType
    -0.68
    20439
    -0.63
     WCS
    -0.63
     unfocusedRange
    -0.59
     CET
    -0.58
     EVs
    -0.58
    Compare
    -0.57
    POSITIVE LOGITS
     disappeared
    1.01
     himself
    0.91
     vanished
    0.90
     testified
    0.88
     allegedly
    0.87
     survived
    0.83
     suffered
    0.82
    's
    0.81
     died
    0.81
     reportedly
    0.80
    Activations Density 0.407%

    No Known Activations