© 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. 22087
    Prev
    Next
    INDEX
    Explanations

    negations and prohibitive statements related to actions or conditions

    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
    53
    +0.10
    0.4%
    567
    +0.10
    0.4%
    101
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    551
    +0.10
    0.04
    643
    +0.10
    0.03
    567
    +0.10
    0.03
    Negative Logits
     Wid
    -0.65
     enhancements
    -0.62
     irresistible
    -0.62
     restoration
    -0.62
     Corona
    -0.61
     Niet
    -0.60
     FW
    -0.60
     Hungry
    -0.59
    ynthesis
    -0.58
     superb
    -0.58
    POSITIVE LOGITS
    inventoryQuantity
    0.76
    erest
    0.75
     underestimated
    0.73
     bother
    0.73
     blindly
    0.72
     EVER
    0.72
     underestimate
    0.69
    reated
    0.69
    roach
    0.69
     necessarily
    0.69
    Activations Density 0.099%

    No Known Activations