© 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. 13277
    Prev
    Next
    INDEX
    Explanations

    information related to healthcare policies and medical conditions, particularly focusing on coverage for pre-existing conditions and premium tax credits based on income levels

    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
    +0.14
    0.7%
    235
    +0.11
    0.5%
    431
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    235
    +0.14
    0.04
    206
    +0.11
    0.03
    710
    +0.09
    0.03
    Negative Logits
    gom
    -0.74
    Prosecut
    -0.74
    wang
    -0.73
    Poké
    -0.71
    ESPN
    -0.69
    abre
    -0.68
    mony
    -0.67
    Serv
    -0.66
    ENGTH
    -0.65
    cade
    -0.65
    POSITIVE LOGITS
     disabilities
    1.08
     poorer
    0.98
     disadvantaged
    0.93
     low
    0.93
     wealthier
    0.93
     deficient
    0.89
     households
    0.87
     poverty
    0.84
     homes
    0.83
     weaker
    0.83
    Activations Density 0.154%

    No Known Activations