© 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. 5303
    Prev
    Next
    INDEX
    Explanations

    words related to challenging situations or conflicts between individuals or groups

    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
    715
    +0.15
    0.7%
    569
    +0.14
    0.6%
    288
    +0.12
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    715
    +0.15
    0.01
    569
    +0.14
    0.01
    536
    +0.12
    0.01
    Negative Logits
     extremes
    -0.70
     Seas
    -0.69
     Sawyer
    -0.66
     Salvation
    -0.66
     grandson
    -0.65
     breath
    -0.65
     mate
    -0.64
     grandfather
    -0.63
    hood
    -0.62
     Slug
    -0.62
    POSITIVE LOGITS
    ilitation
    1.34
    ulty
    1.26
    ilit
    1.17
    eless
    1.16
    etime
    1.15
    ility
    1.12
    ep
    1.11
    ilities
    1.10
    sim
    1.08
    ework
    1.04
    Activations Density 0.019%

    No Known Activations