© 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. 17251
    Prev
    Next
    INDEX
    Explanations

    instances where someone is taking over responsibilities or roles from another person

    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
    505
    +0.11
    0.5%
    153
    +0.10
    0.5%
    264
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    505
    +0.11
    0.06
    526
    +0.10
    0.00
    554
    +0.10
    0.00
    Negative Logits
    raq
    -0.72
    apa
    -0.68
    ileaks
    -0.65
     Prices
    -0.64
    heres
    -0.63
    hend
    -0.63
    Methods
    -0.61
    htaking
    -0.61
    bidden
    -0.60
    ochet
    -0.60
    POSITIVE LOGITS
     role
    1.55
     roles
    1.49
    role
    1.32
     duties
    1.31
     slot
    1.22
     position
    1.17
     mantle
    1.13
     helm
    1.12
     throne
    1.11
     responsibilities
    1.06
    Activations Density 0.554%

    No Known Activations