© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 33654
    Prev
    Next
    INDEX
    Explanations

    assignment, variable declarations

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

    The neuron flags tokens that appear in code conditional expressions—especially parentheses and comparison‐or‐assignment operators (e.g. “while((…)=…)!=…” and “false !== ($…)”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     både
    -1.12
    naa
    -1.09
    upb
    -1.05
     chcia
    -1.03
     requieren
    -1.01
    
    -1.01
     empiezan
    -1.00
     différentes
    -0.98
    Knows
    -0.98
     charlotte
    -0.98
    POSITIVE LOGITS
     by
    1.29
     from
    1.28
     one
    1.04
     at
    1.04
     because
    0.98
     through
    0.94
     on
    0.93
    __);
    0.91
    thirds
    0.91
     after
    0.91
    Activations Density 0.002%

    No Known Activations