© 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. 13282
    Prev
    Next
    INDEX
    Explanations

    industrial and governmental terms related to nationalization and legal disputes

    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
    +0.12
    0.5%
    151
    +0.12
    0.5%
    224
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.12
    0.07
    146
    +0.12
    0.07
    288
    +0.12
    0.08
    Negative Logits
     intuition
    -0.65
    GoldMagikarp
    -0.64
     Jackets
    -0.63
    Proof
    -0.61
     pecul
    -0.61
    keyes
    -0.61
     increments
    -0.61
    uristic
    -0.60
     defin
    -0.59
    ipeg
    -0.59
    POSITIVE LOGITS
     alleging
    0.88
     protesting
    0.86
     amid
    0.83
     whom
    0.79
    interstitial
    0.78
     Corinth
    0.73
     unfairly
    0.72
     discriminated
    0.71
     hostage
    0.71
     clashed
    0.71
    Activations Density 0.423%

    No Known Activations