© 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. 1440
    Prev
    Next
    INDEX
    Explanations

    news about criminal activities

    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
    629
    +0.13
    0.6%
    509
    +0.10
    0.5%
    175
    +0.10
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    509
    +0.13
    0.05
    204
    +0.10
    0.05
    629
    +0.10
    0.04
    Negative Logits
     boycot
    -0.82
     losers
    -0.79
    puff
    -0.77
     impover
    -0.75
    etus
    -0.74
     starving
    -0.73
     sake
    -0.72
    Tea
    -0.72
    ayn
    -0.72
     boycott
    -0.71
    POSITIVE LOGITS
     Detect
    1.35
     Detective
    1.27
     detectives
    1.26
     Forensic
    1.24
     Enforcement
    1.24
     Investigators
    1.17
     Police
    1.17
     Officer
    1.17
    police
    1.17
    intendent
    1.16
    Activations Density 0.512%

    No Known Activations