© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 44799
    Prev
    Next
    INDEX
    Explanations

    mathematical variables or translocations

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

    The neuron is tuned to pick out the variable “t” (often in exponentiation, like t**2 or t**3) within mathematical or code‐like expressions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    occurrences of the variable “t” and t-prefixed identifiers, especially in algebraic expressions and technical/code notation.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the variable or token "t" in mathematical expressions and code.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
    m
    -2.03
    ↵
    -1.99
     or
    -1.84
    Tôi
    -1.63
    s
    -1.63
    úan
    -1.59
    n
    -1.56
    z
    -1.52
    p
    -1.51
    ftrag
    -1.46
    POSITIVE LOGITS
     payé
    1.86
    籥
    1.80
     tenté
    1.67
     attendu
    1.66
     apporté
    1.63
    
    1.60
    咴
    1.57
    ruguay
    1.54
    aile
    1.54
     joué
    1.53
    Activations Density 0.003%

    No Known Activations