© 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. 34-GEMMASCOPE-RES-131K
    7. 117175
    Prev
    Next
    INDEX
    Explanations

    skincare types

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

    descriptions of skin and hair products, particularly focusing on dry, damaged, or problem conditions.

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

    The neuron fires strongly on “for” phrases in product–marketing text—that is, tokens that introduce the intended use or benefit (e.g. “for dry, damaged, tangled hair”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     Fluids
    -0.90
    🐶
    -0.83
    CEED
    -0.83
    Marital
    -0.82
    ceptor
    -0.82
     warmth
    -0.80
    неры
    -0.79
    頼
    -0.79
    vernment
    -0.78
     тепло
    -0.78
    POSITIVE LOGITS
     Type
    1.43
     type
    1.43
    Type
    1.23
     Types
    1.14
     types
    1.14
     normal
    1.13
     TYPE
    1.12
    Тип
    1.10
    getType
    1.09
    TYPE
    1.09
    Activations Density 0.024%

    No Known Activations