© 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. 34-GEMMASCOPE-RES-131K
    7. 75543
    Prev
    Next
    INDEX
    Explanations

    breaking and entering

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

    The neuron detects language describing unauthorized entry or breaking and entering (e.g. intruders, burglary, entering through open windows).

    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
     clutches
    -0.78
    фу
    -0.74
     sting
    -0.74
    klär
    -0.73
    zyp
    -0.72
    bitField
    -0.72
    geln
    -0.69
    zten
    -0.68
    чку
    -0.68
     confis
    -0.68
    POSITIVE LOGITS
     breaking
    1.79
     break
    1.63
     broke
    1.49
     forcibly
    1.49
     forced
    1.48
    breaking
    1.44
     force
    1.42
    Breaking
    1.41
    break
    1.38
     Breaking
    1.37
    Activations Density 0.034%

    No Known Activations