© 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. 40157
    Prev
    Next
    INDEX
    Explanations

    disabilities after tissue

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

    punctuation marks, especially commas, at the end of sentences or clauses.

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

    The neuron is acting as a positional‐encoding detector, firing more strongly on tokens that occur later (farther along) in the input sequence.

    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
    ipts
    -1.09
     сочета
    -1.09
    ACEAE
    -1.00
    CLES
    -0.99
    也不能
    -0.98
     gewährle
    -0.98
     tästä
    -0.97
     manifestaciones
    -0.96
    人也
    -0.96
     valamint
    -0.96
    POSITIVE LOGITS
     when
    1.21
     to
    1.12
     because
    0.97
     will
    0.95
    CONCLUSIONES
    0.93
    芣
    0.93
     by
    0.93
    -
    0.92
    DetailActivity
    0.91
     большим
    0.89
    Activations Density 0.064%

    No Known Activations