© 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. 98557
    Prev
    Next
    INDEX
    Explanations

    rounding and listing

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

    The neuron specializes in spotting rounding instructions—words like “round(ed),” “to,” “nearest,” “decimal,” “place(s),” “dps,” and “integer” that specify how many decimal places or to what unit a number should be rounded.

    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
     stiller
    -1.10
     keny
    -1.06
     Ráp
    -1.04
     ſome
    -1.03
     certains
    -1.02
     kategor
    -1.01
     Fácil
    -1.01
     elástico
    -1.00
     métiers
    -1.00
    ſen
    -0.99
    POSITIVE LOGITS
     with
    1.23
     때
    1.09
     if
    1.07
    utriche
    1.03
     using
    1.00
     TWO
    1.00
    arthed
    1.00
    чя
    0.99
     in
    0.96
     також
    0.96
    Activations Density 0.031%

    No Known Activations