© 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. 43339
    Prev
    Next
    INDEX
    Explanations

    conditioned or conditioning

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

    The neuron is specialized to detect occurrences of the root “condition,” i.e. words like conditioning, conditioned, preconditioning, 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
    org
    -1.40
    })-
    -1.38
     hade
    -1.38
     económico
    -1.38
    ;:
    -1.36
    想到了
    -1.35
    起码
    -1.34
     leaft
    -1.32
    就没
    -1.32
    ?-
    -1.32
    POSITIVE LOGITS
    Since
    1.68
    Also
    1.67
    According
    1.57
    Each
    1.55
    Although
    1.53
     这个
    1.53
    These
    1.49
     consentement
    1.46
     років
    1.44
    There
    1.35
    Activations Density 0.087%

    No Known Activations