© 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. 111053
    Prev
    Next
    INDEX
    Explanations

    it must be

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

    The neuron strongly activates on short pronoun tokens—especially “it,” “there,” “you,” etc.—i.e. it’s detecting subject/expletive pronouns.

    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
     devraient
    -1.02
    Need
    -1.01
     Need
    -0.91
    laires
    -0.88
     стоит
    -0.85
    Needs
    -0.85
     besoin
    -0.81
     tarvit
    -0.81
    tigheter
    -0.79
     Needs
    -0.77
    POSITIVE LOGITS
     must
    7.47
    must
    5.13
    Must
    5.03
     Must
    4.41
     måste
    3.33
     MUST
    3.14
    MUST
    3.06
    必须
    3.05
     doivent
    2.94
     debe
    2.89
    Activations Density 0.150%

    No Known Activations