© 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. 34-GEMMASCOPE-RES-131K
    7. 108264
    Prev
    Next
    INDEX
    Explanations

    pity and sympathy

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

    The neuron is especially sensitive to very common function words (pronouns, articles, prepositions, auxiliaries) rather than content words.

    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
    ření
    -0.93
    ᥙ
    -0.87
    贄
    -0.79
    ppb
    -0.77
    łaj
    -0.77
     Lb
    -0.76
    PSP
    -0.75
    μον
    -0.74
    ifera
    -0.74
     bonheur
    -0.73
    POSITIVE LOGITS
     pity
    5.41
     sympathy
    4.28
    pity
    3.48
     Pity
    3.47
     compassion
    3.20
     sympathetic
    2.98
     sympathi
    2.94
     sympathize
    2.92
     empathy
    2.84
    symp
    2.84
    Activations Density 0.045%

    No Known Activations