© 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. 3176
    Prev
    Next
    INDEX
    Explanations

    crew workmanship move

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

    The neuron detects highly positive, superlative praise or endorsement language (e.g. “100 % satisfied,” “absolutely,” “would recommend”) in customer-review style text.

    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
     befinden
    -0.92
     meraih
    -0.86
    häuse
    -0.85
    -0.83
    eloma
    -0.83
                 
    -0.82
    makeText
    -0.81
              
    -0.81
    ɹ
    -0.81
     pertinentes
    -0.81
    POSITIVE LOGITS
     crew
    1.27
     They
    1.16
     workmen
    1.12
     foreman
    1.10
     they
    1.09
     estimate
    1.06
     subcontra
    1.06
     subcontractors
    1.05
     workers
    1.02
     communicated
    1.00
    Activations Density 0.011%

    No Known Activations