© 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. 32836
    Prev
    Next
    INDEX
    Explanations

    human worldly carnal nature

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

    The neuron fires on contrastive or concessive discourse markers (e.g. “not…,” “but,” “even when,” “don’t,” “when”) that signal a reversal or exception.

    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
     胶
    -0.89
     Moines
    -0.83
    tilo
    -0.78
    Getenv
    -0.76
     Studios
    -0.71
     julka
    -0.71
    ǘ
    -0.70
    ğaz
    -0.70
     Versch
    -0.69
     院
    -0.69
    POSITIVE LOGITS
     human
    1.77
    human
    1.53
     earthly
    1.31
    Human
    1.28
     Human
    1.20
     menschliche
    1.16
     relying
    1.16
     manusia
    1.15
     worldly
    1.13
     челове
    1.10
    Activations Density 0.018%

    No Known Activations