© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 128872
    Prev
    Next
    INDEX
    Explanations

    again, issuance, profile, shedding

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

    The neuron is detecting verb-particle constructions (phrasal verbs and similar verb + adverbial particle combinations, e.g. “look out,” “works together,” “all over”).

    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
    political
    -1.51
     about
    -1.32
     much
    -1.23
     ease
    -1.23
     they
    -1.17
     different
    -1.16
    different
    -1.14
    SIDE
    -1.13
     support
    -1.13
     dernière
    -1.11
    POSITIVE LOGITS
     yine
    1.80
    那些
    1.65
     AGAIN
    1.61
    );
    1.60
    ៖
    1.52
    Again
    1.52
     again
    1.48
     amarillas
    1.48
     potrebné
    1.44
    '''
    1.43
    Activations Density 0.165%

    No Known Activations