© 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
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 8728
    Prev
    Next
    INDEX
    Explanations

    not alone

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

    The neuron flags first-person help/questions—i.e. expressions of “I” asking if something is missing or offering to provide more information.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    the first-person pronoun "I" in informal, conversational contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    first-person singular expressions indicating the speaker’s own actions, opinions, or experiences.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     menghindari
    0.76
     dubious
    0.70
     contempt
    0.70
     versucht
    0.66
     disdain
    0.64
    试图
    0.63
     intento
    0.63
     deliberate
    0.63
     truth
    0.62
    避免
    0.61
    POSITIVE LOGITS
    Wrong
    1.00
    wrong
    0.97
     wrong
    0.91
     Wrong
    0.80
     गलत
    0.73
    错了
    0.73
    Salah
    0.73
     salah
    0.71
    間違
    0.71
    Imag
    0.71
    Activations Density 0.206%

    No Known Activations