© 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. 72790
    Prev
    Next
    INDEX
    Explanations

    correcting errors and alternatives

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

    This neuron detects words and phrases in errata or correction contexts, flagging when a statement is being revised or corrected.

    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
    ฐาน
    -0.92
    resar
    -0.91
    Hlav
    -0.81
    тельными
    -0.81
     devra
    -0.80
    Why
    -0.77
    Who
    -0.77
    etzal
    -0.76
     اض
    -0.75
    észetes
    -0.75
    POSITIVE LOGITS
    而非
    0.94
     instead
    0.94
     correctly
    0.93
     enclosures
    0.91
     corect
    0.90
     correct
    0.89
     XXXVI
    0.88
     invece
    0.88
     correctamente
    0.85
    じゃなくて
    0.82
    Activations Density 0.045%

    No Known Activations