© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 98219
    Prev
    Next
    INDEX
    Explanations

    introducing reasons

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

    The neuron strongly activates on causal connectives and their surrounding tokens (e.g. “because,” “coz,” “it,” “will”)—in other words, it detects reason-giving or explanatory clauses.

    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
     chistes
    -1.23
    baliknya
    -1.22
     retratos
    -1.18
     hudeb
    -1.17
     recortes
    -1.13
    İyi
    -1.12
     repuesto
    -1.12
     esquí
    -1.10
    بغ
    -1.09
     vám
    -1.08
    POSITIVE LOGITS
     recently
    1.44
     previously
    1.27
     they
    1.27
     witnessed
    1.16
     wcześniej
    1.15
     its
    1.13
    其
    1.12
     joão
    1.11
     majority
    1.08
    Its
    1.08
    Activations Density 0.033%

    No Known Activations