© 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. 6916
    Prev
    Next
    INDEX
    Explanations

    words related to violent or dangerous situations

    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
    373
    +0.12
    0.6%
    659
    +0.11
    0.5%
    620
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    284
    +0.12
    0.03
    659
    +0.11
    0.02
    598
    +0.10
    0.02
    Negative Logits
     cowork
    -0.69
    =-=-
    -0.65
     Presbyter
    -0.59
     toget
    -0.59
     initials
    -0.58
     loyalty
    -0.57
    tons
    -0.57
     rift
    -0.57
     rides
    -0.57
     Payton
    -0.57
    POSITIVE LOGITS
    pport
    1.06
    cius
    1.03
    cca
    1.01
    oti
    0.98
    endo
    0.98
    asion
    0.95
    arant
    0.94
    ccess
    0.94
    pta
    0.93
    ña
    0.92
    Activations Density 0.025%

    No Known Activations