© 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. 1187
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on domain-specific technical nouns (e.g. object/field identifiers in code docs, tool and hardware names, material or device terms in product or scientific descriptions).

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

    technical or product-related nouns in descriptive or documentation contexts.

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

    mentions of concrete, countable nouns referring to specific physical objects or items, especially technical or product-related entities.

    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
    বর্তন
    0.66
     помогает
    0.65
    のために
    0.65
    Responsible
    0.63
     enables
    0.62
    Helping
    0.61
    ിലേക്ക്
    0.61
    是为了
    0.60
    Inside
    0.58
    箇所
    0.57
    POSITIVE LOGITS
     whose
    1.43
     with
    1.42
     containing
    1.32
     עם
    1.32
    ที่มี
    1.30
     cuyas
    1.27
     dengan
    1.24
     với
    1.24
     cuyo
    1.24
     със
    1.24
    Activations Density 1.029%

    No Known Activations