© 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. 22-GEMMASCOPE-RES-131K
    7. 34043
    Prev
    Next
    INDEX
    Explanations

    possessive determiners introducing a core purpose

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

    The neuron detects words that signal personal or collective ownership and emphasize primacy—i.e. possessive pronouns and priority‐marking adjectives.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    zös
    -1.30
    utando
    -1.19
     onCancelled
    -1.14
     menyen
    -1.14
    orz
    -1.13
     dakika
    -1.13
     vokal
    -1.12
     հղ
    -1.11
     gitar
    -1.10
    llte
    -1.09
    POSITIVE LOGITS
     goal
    2.00
     aim
    1.87
     goals
    1.53
     purpose
    1.53
     objective
    1.49
     objectives
    1.48
     strategy
    1.46
     aims
    1.43
     intent
    1.41
     focus
    1.34
    Activations Density 0.038%

    No Known Activations