© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 24200
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate competition or opposition

    oai_token-act-pair · gpt-4o-miniTriggered 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
    231
    +0.10
    0.5%
    238
    +0.10
    0.5%
    496
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    305
    +0.10
    0.03
    238
    +0.10
    0.02
    281
    +0.09
    0.02
    Negative Logits
    pmwiki
    -0.81
    oola
    -0.77
    aund
    -0.73
    milo
    -0.73
    yright
    -0.71
    ortion
    -0.71
    tube
    -0.70
    verified
    -0.70
    udder
    -0.70
    iscal
    -0.69
    POSITIVE LOGITS
     opponents
    0.88
     opponent
    0.81
     Notre
    0.81
     adversity
    0.81
     Samoa
    0.80
     foes
    0.80
     Nicaragua
    0.77
     Enemy
    0.75
     Suzuki
    0.74
     Carolina
    0.72
    Activations Density 0.039%

    No Known Activations