© 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. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 60182
    Prev
    Next
    INDEX
    Explanations

    criminal charges, court, damage, use

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron detects mentions of criminal‐justice language—especially the word “criminal” (and related legal terms).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    还是要
    -1.06
    不是很
    -0.96
    そこまで
    -0.87
     sooo
    -0.86
    \
    -0.85
    還好
    -0.85
     tumultuous
    -0.84
    甭
    -0.83
    就没
    -0.83
     soooo
    -0.82
    POSITIVE LOGITS
     enterprise
    1.26
     mastermind
    1.23
     Minds
    1.22
     enterprises
    1.18
    enterprise
    1.16
     elements
    1.10
     element
    1.09
     minds
    1.06
     justice
    1.02
     activity
    0.99
    Activations Density 0.014%

    No Known Activations