© 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. 17232
    Prev
    Next
    INDEX
    Explanations

    news events and updates related to specific locations or incidents

    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
    373
    +0.15
    0.7%
    330
    +0.12
    0.6%
    136
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    388
    +0.15
    0.02
    107
    +0.12
    0.02
    373
    +0.09
    -0.01
    Negative Logits
     Telecommunications
    -0.66
     spawns
    -0.63
     DRAGON
    -0.63
     shelters
    -0.61
     respawn
    -0.59
     Guardians
    -0.59
     Utilities
    -0.59
     Forestry
    -0.58
     summons
    -0.57
     Warfare
    -0.57
    POSITIVE LOGITS
    aspx
    0.91
    dp
    0.84
    html
    0.83
    /?
    0.77
    tnc
    0.75
    201
    0.73
    320
    0.73
    comment
    0.72
    src
    0.71
    471
    0.71
    Activations Density 0.047%

    No Known Activations