© 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. 112614
    Prev
    Next
    INDEX
    Explanations

    told [recipient]

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

    The neuron detects occurrences of “told me,” i.e. first-person reports of someone having told the speaker something.

    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
    ２
    -2.53
    what
    -2.41
     resolute
    -2.36
     unwavering
    -2.34
    ities
    -2.33
    １
    -2.33
     на
    -2.31
    -"
    -2.30
    BeNil
    -2.28
     ''
    -2.23
    POSITIVE LOGITS
    
    3.56
    猡
    3.50
     inverno
    2.98
    饪
    2.98
    僦
    2.91
    他们在
    2.83
     ポスター
    2.72
    痦
    2.72
    豋
    2.69
     isolamento
    2.61
    Activations Density 0.007%

    No Known Activations