© 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. 35874
    Prev
    Next
    INDEX
    Explanations

    words ending in 'constructor' or 'ally'

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

    The neuron strongly activates on specialized or domain-specific terminology (e.g. technical nouns and jargon).

    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
     Resmi
    -1.05
     asio
    -1.03
     lyder
    -0.97
     Sosial
    -0.94
    过来的
    -0.93
    詫
    -0.92
    餃子
    -0.91
    过来了
    -0.91
    出来了
    -0.91
     demselben
    -0.91
    POSITIVE LOGITS
     Hinsicht
    1.04
     veröffentlich
    0.98
     construite
    0.96
    に行
    0.95
     žaid
    0.93
     preocupaciones
    0.93
    éről
    0.92
    vět
    0.91
    vyš
    0.90
     posibilidades
    0.89
    Activations Density 0.052%

    No Known Activations