© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 9489
    Prev
    Next
    INDEX
    Explanations

    introduces observation

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

    The neuron detects second‐person perception or evidence cues—phrases like “you can see,” “as you can see,” and “view”—that point the reader to look at or notice something.

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

    phrases introducing visual or observational statements, particularly "As you can see" and similar constructions that direct the reader's attention.

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

    phrases that directly address the reader to point out or highlight something, typically using second-person perception verbs like seeing or noticing.

    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_24_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
     mejor
    0.58
    ynchronously
    0.58
     meglio
    0.57
     analyte
    0.55
     intellig
    0.55
     sensorial
    0.54
     consid
    0.54
     mieux
    0.54
     transceiver
    0.54
    ならば
    0.54
    POSITIVE LOGITS
     Doesn
    0.84
     शून्य
    0.80
     就是
    0.80
     जीरो
    0.76
    Doesn
    0.76
     Están
    0.76
     없습니다
    0.75
     Today
    0.74
     Bugün
    0.74
     сейчас
    0.73
    Activations Density 0.013%

    No Known Activations