© 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. 34-GEMMASCOPE-RES-131K
    7. 103181
    Prev
    Next
    INDEX
    Explanations

    something holds true

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

    The neuron activates on explicit statements of mathematical inequalities—specifically on occurrences of the word “inequality” (often paired with “true”) in a technical context.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    🎚
    -0.91
    Ⓟ
    -0.81
    zené
    -0.81
    υν
    -0.80
    Buna
    -0.77
    )++;
    -0.77
    آموزش
    -0.76
    зно
    -0.75
    magne
    -0.75
    hirkan
    -0.75
    POSITIVE LOGITS
     holds
    3.53
     hold
    3.02
    holds
    2.53
     Holds
    2.45
     holding
    2.38
     Hold
    2.20
    hold
    2.19
    Hold
    2.13
     held
    2.08
    HOLD
    1.82
    Activations Density 0.063%

    No Known Activations