© 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
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72588
    Prev
    Next
    INDEX
    Explanations

    exponent and scaling

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

    The neuron specializes in spotting technical scientific expressions of quantitative relationships—especially mentions of power‐law behavior, exponents, distributions, scaling laws, and other numeric parameter terms.

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

    scientific or mathematical text describing power law relationships, exponents, and scaling behavior.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    scientific descriptions of scale-invariant phenomena quantified by exponents and universality properties.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    .
    -1.02
    Consider
    -0.95
     perigo
    -0.94
     brilhante
    -0.91
     '~
    -0.90
     două
    -0.88
     espelho
    -0.86
    さの
    -0.86
     圣
    -0.85
     rca
    -0.85
    POSITIVE LOGITS
    requipa
    1.29
     for
    1.16
     नीचे
    1.16
    Vaata
    1.13
    独特
    1.09
    łaści
    1.08
     través
    1.06
    umenten
    1.05
    styrelsen
    1.05
     atriz
    1.05
    Activations Density 0.027%

    No Known Activations