© 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. 95703
    Prev
    Next
    INDEX
    Explanations

    How to perform an action

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

    The neuron fires strongly on the main instructional action words (title‐cased verbs) used in “How to…” headings or similar tutorial titles.

    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
     phénom
    -1.47
     leopardo
    -1.44
     ſou
    -1.44
     accepte
    -1.43
     Organizador
    -1.39
    乁
    -1.38
     Gerente
    -1.38
     tortuga
    -1.37
     raiſ
    -1.37
     portefeuille
    -1.37
    POSITIVE LOGITS
    </h1>
    2.80
    </h2>
    2.63
    </strong>
    2.28
    </h4>
    2.27
    </h3>
    2.17
    </b>
    2.16
    </h5>
    1.93
     to
    1.83
     How
    1.78
     using
    1.76
    Activations Density 0.078%

    No Known Activations