© 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. 86421
    Prev
    Next
    INDEX
    Explanations

    Prakash Raj, Mubarak, RDP

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

    The neuron is essentially detecting the letter‐sequence “rak” (in any context or casing) inside individual tokens.

    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
    );
    -3.00
    ↵
    -2.92
    ),
    -2.78
    ',
    -2.42
     GOtt
    -2.28
    that
    -2.28
     tht
    -2.20
     neue
    -2.11
     With
    -2.08
    ！）
    -2.03
    POSITIVE LOGITS
    geſ
    2.78
    ਞ
    2.63
    ቍ
    2.53
    穑
    2.48
    ta
    2.48
    錞
    2.45
    ureusement
    2.33
    谖
    2.31
    棪
    2.31
    らの
    2.30
    Activations Density 0.004%

    No Known Activations