© 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. 22-GEMMASCOPE-RES-131K
    7. 15106
    Prev
    Next
    INDEX
    Explanations

    niobium oxide, WOLEDs, based on

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

    scientific material terms and elements related to batteries, solar cells, and electronic components.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    This neuron responds to dense technical language and domain-specific jargon in scientific and engineering contexts—especially materials-science terms and property descriptions in research abstracts.

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

    technical terminology in scientific and academic texts about materials science, batteries, solar cells, and electronic devices.

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

    technical materials-science language about energy conversion and storage devices (e.g., batteries, solar cells, LEDs, membranes, thermoelectrics) and their performance descriptions.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     its
    -1.52
     it
    -1.27
     their
    -1.25
    创建时间
    -1.24
    ViewFeatures
    -1.22
     угло
    -1.21
    Trọng
    -1.21
     baloncesto
    -1.20
    InstrumentedTest
    -1.20
    цыі
    -1.16
    POSITIVE LOGITS
     rootReducer
    1.38
     教程
    1.27
     candidate
    1.25
    quê
    1.17
     简介
    1.14
    lampu
    1.14
     tien
    1.14
     сахара
    1.09
    coba
    1.08
     uda
    1.08
    Activations Density 0.199%

    No Known Activations