© 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. 2428
    Prev
    Next
    INDEX
    Explanations

    phrases related to the concept of impermanence

    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
    308
    +0.11
    0.5%
    59
    +0.10
    0.5%
    232
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.11
    0.01
    503
    +0.10
    0.01
    549
    +0.10
    0.01
    Negative Logits
    anwhile
    -0.78
    wagen
    -0.71
    gdala
    -0.69
     Ou
    -0.64
    FactoryReloaded
    -0.64
    âĸ¬
    -0.64
    OCK
    -0.61
     Hath
    -0.61
     Highlander
    -0.59
     deductions
    -0.59
    POSITIVE LOGITS
    manent
    1.36
    vious
    1.33
    missible
    1.23
    ishable
    1.14
    bably
    0.93
    cept
    0.93
    lesi
    0.91
    ceivable
    0.91
    ensible
    0.90
    miss
    0.90
    Activations Density 0.011%

    No Known Activations