© 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. 10-GEMMASCOPE-RES-131K
    7. 104823
    Prev
    Next
    INDEX
    Explanations

    act quickly, following tests, after being

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

    The neuron flags first‐person confessional or apology language, e.g. personal admissions of poor judgment or misconduct.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    directly quoted, formal statements—press-release or spokesperson/official remarks enclosed in quotation marks.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    formal statements or announcements, particularly those using diplomatic or corporate language to explain decisions, actions, or policies.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
     fraîche
    -1.43
     sembla
    -1.41
     fraiche
    -1.37
     chaude
    -1.31
    ぼり
    -1.24
    膂
    -1.23
     tropicales
    -1.23
     brazos
    -1.23
     getLayout
    -1.23
    ろし
    -1.22
    POSITIVE LOGITS
     your
    1.75
     its
    1.63
     or
    1.43
     will
    1.40
     probablemente
    1.37
     onderwerp
    1.30
    ITECH
    1.28
    始
    1.28
    ></
    1.27
    vyšší
    1.23
    Activations Density 0.101%

    No Known Activations