© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 54089
    Prev
    Next
    INDEX
    Explanations

    as you always do

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

    The neuron fires on words in that specific post‐game quote about pre-game planning—i.e. it detects talk of “plan[ning] everything during the week, but you don’t plan to lose a man five minutes into the game.”

    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
     jeszcze
    -1.03
    ))=
    -0.93
     extraordinaire
    -0.92
    า
    -0.91
     tampil
    -0.91
    "]
    -0.90
     yet
    -0.90
     iLocal
    -0.88
    lawan
    -0.88
     joueur
    -0.88
    POSITIVE LOGITS
     always
    1.54
     happens
    1.32
    這種
    1.25
     ilyen
    1.22
     such
    1.20
    就是这样
    1.17
    Such
    1.16
     таких
    1.16
     solche
    1.14
     завжди
    1.14
    Activations Density 0.011%

    No Known Activations