© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 128923
    Prev
    Next
    INDEX
    Explanations

    introducing conclusive statements

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

    the presence of sentence-level transition words or conjunctive adverbs (e.g. “Therefore,” “Hence,” “Accordingly,” “Thus,” etc.).

    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
    辶
    -1.51
    Contrato
    -1.45
    ↵↵
    -1.42
    ла
    -1.40
     solch
    -1.39
     både
    -1.39
    Характеристики
    -1.39
    ⦊
    -1.38
     such
    -1.36
    CHECKBOX
    -1.36
    POSITIVE LOGITS
    ,
    2.00
    "
    1.64
    难怪
    1.60
    0
    1.55
    1.53
    _
    1.46
    倘
    1.45
    ?
    1.42
    非常好
    1.41
    em
    1.39
    Activations Density 0.164%

    No Known Activations