© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 60667
    Prev
    Next
    INDEX
    Explanations

    medical and technical terms

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

    character patterns in regular expressions, particularly numeric ranges like [0-9].

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

    The main thing this neuron does is detect numeric characters (digits) in code or regular‐expression contexts.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     Each
    -1.38
    there
    -1.34
    അ
    -1.32
     Geschä
    -1.32
     ecosystems
    -1.31
     infamous
    -1.30
     dieną
    -1.28
     each
    -1.27
    釭
    -1.26
    輜
    -1.25
    POSITIVE LOGITS
    艄
    1.59
    lysm
    1.59
    gonna
    1.58
    缡
    1.51
     мужская
    1.50
    干什么
    1.41
    fahan
    1.41
    gotta
    1.39
    égio
    1.39
    causes
    1.37
    Activations Density 0.006%

    No Known Activations