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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 24192
    Prev
    Next
    INDEX
    Explanations

    terms related to health, well-being, and improvement outcomes

    oai_token-act-pair · gpt-4o-miniTriggered 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
    481
    +0.18
    0.9%
    447
    +0.14
    0.6%
    720
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    393
    +0.18
    0.05
    285
    +0.14
    0.04
    225
    +0.11
    0.05
    Negative Logits
    quished
    -0.92
    wic
    -0.85
     Cosponsors
    -0.85
    ctuary
    -0.81
    utory
    -0.80
    ulhu
    -0.76
    oops
    -0.71
    ombies
    -0.70
    letter
    -0.70
    ãĥ¼ãĥ³
    -0.68
    POSITIVE LOGITS
     responsiveness
    1.32
     productivity
    1.32
     competitiveness
    1.28
     mobility
    1.27
     susceptibility
    1.25
     longevity
    1.25
     viability
    1.25
     attractiveness
    1.25
     wellbeing
    1.20
     effectiveness
    1.20
    Activations Density 0.235%

    No Known Activations