© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 13215
    Prev
    Next
    INDEX
    Explanations

    still / also followed by descriptive words

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

    The neuron fires on common discourse‐marking constructions around the adverb “still” (e.g. “it is still important to…,” “there are still some…”) that signal continuation or emphasis.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     but
    -2.05
     though
    -1.88
     when
    -1.79
     while
    -1.78
     although
    -1.70
     nhưng
    -1.64
     ولكن
    -1.55
     però
    -1.55
     But
    -1.52
     zwar
    -1.52
    POSITIVE LOGITS
     compris
    1.43
    …»
    1.38
     dalamnya
    1.38
    着一个
    1.36
     ڤ
    1.34
    olving
    1.28
    ఝ
    1.27
    ...»
    1.27
     ?",
    1.27
    descri
    1.27
    Activations Density 0.172%

    No Known Activations