© 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. 22-GEMMASCOPE-RES-131K
    7. 71038
    Prev
    Next
    INDEX
    Explanations

    step after step

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

    The neuron fires on procedural “how-to” instruction language—words that mark steps or actions in tutorials (e.g. off, with, so, must, will, install, first, then).

    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
     other
    -1.22
     quando
    -1.20
     otros
    -1.09
     even
    -1.06
     cuando
    -1.05
    上げた
    -1.04
    rašymas
    -1.03
    其他
    -0.99
    甚至
    -0.99
    ⽼
    -0.98
    POSITIVE LOGITS
     again
    1.34
    again
    1.30
     by
    1.25
     dilakukan
    1.09
    Again
    1.08
    これも
    1.08
     novamente
    1.05
     opět
    1.04
     Again
    1.03
     musíte
    1.03
    Activations Density 0.036%

    No Known Activations