© 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. 20379
    Prev
    Next
    INDEX
    Explanations

    terms related to political controversies and violations, such as war crimes, labor exploitation, racism, and torture

    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
    481
    +0.19
    0.9%
    447
    +0.14
    0.7%
    387
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.19
    0.08
    439
    +0.14
    0.04
    126
    +0.10
    0.04
    Negative Logits
    utterstock
    -0.69
     wisely
    -0.67
    adle
    -0.65
     reins
    -0.61
     Comedy
    -0.60
    ç¥ŀ
    -0.60
    agonist
    -0.59
    APTER
    -0.59
     STATS
    -0.59
     Conrad
    -0.58
    POSITIVE LOGITS
     irregularities
    1.11
     discrepancies
    0.99
     wrongdoing
    0.97
    Semitism
    0.97
     contamination
    0.93
     tampering
    0.91
     imminent
    0.90
     impending
    0.89
     involvement
    0.89
     leakage
    0.89
    Activations Density 0.409%

    No Known Activations