© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 40280
    Prev
    Next
    INDEX
    Explanations

    biological, financial, or historical contexts

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

    noun-noun compound pairs or adjective-noun pairs that form technical terms.

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

    The neuron fires on individual content-bearing nouns or short technical terms—especially the key words at the start of multiword noun phrases (e.g. “basic position,” “electricity bills,” “isolated ITx,” “acetylcholine receptors,” “package.json,” etc.).

    oai_token-act-pair · o4-miniTriggered 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
    になれば
    -1.38
     this
    -1.33
    為に
    -1.25
     their
    -1.12
    来週
    -1.09
    ほら
    -1.05
    を変更
    -1.01
     tudi
    -1.01
    ところに
    -0.99
    Então
    -0.98
    POSITIVE LOGITS
    thodoxy
    1.25
    urals
    1.11
     การ
    1.10
    phans
    1.05
    bores
    1.05
    antiate
    1.04
     pne
    1.03
     fleurs
    1.03
    😞
    1.02
     and
    1.01
    Activations Density 0.079%

    No Known Activations