© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 6126
    Prev
    Next
    INDEX
    Explanations

    The neuron detects section-header keywords that introduce structured information subsections—e.g. “features,” “specifications,” “reviews,” “guarantees,” etc.

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

    product descriptions and specifications in e-commerce or catalog listings.

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

    structured, catalog-like sections that enumerate detailed specifications, features, descriptions, and other metadata for products, places, or reports.

    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
     azi
    0.61
     ：
    0.59
    ：
    0.59
    :
    0.58
    ????????
    0.58
    最多
    0.57
    :%
    0.57
     CRF
    0.57
    uffed
    0.57
    ：%
    0.56
    POSITIVE LOGITS
     Remarks
    1.09
    備考
    1.07
     특징
    1.04
     Limitations
    0.98
     Details
    0.98
     Constraints
    0.98
     Notes
    0.96
    注意事项
    0.95
     Features
    0.94
     বৈশিষ্ট্য
    0.94
    Activations Density 0.190%

    No Known Activations