© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 49152
    9. 8-RES_FS49152-JB
    11. 25989
    Prev
    Next
    INDEX
    Explanations

    phrases related to violence and social injustices

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs49152-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    49,152
    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
    Neuron Alignment
    Index
    Value
    % of L₁
    480
    +0.25
    1.2%
    762
    +0.11
    0.5%
    388
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    762
    +0.25
    0.07
    225
    +0.11
    0.05
    132
    +0.10
    0.04
    Negative Logits
    osite
    -0.64
    CLASSIFIED
    -0.61
    ocry
    -0.58
    endum
    -0.57
    trak
    -0.57
    ongo
    -0.56
     Outlook
    -0.56
    ANC
    -0.55
    Reloaded
    -0.55
    ortium
    -0.55
    POSITIVE LOGITS
     anymore
    0.87
     instead
    0.84
     whilst
    0.81
     blah
    0.79
     while
    0.78
    .;
    0.78
     lest
    0.77
     outweigh
    0.77
     ruining
    0.75
    .#
    0.75
    Activations Density 0.366%

    No Known Activations