© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 39310
    Prev
    Next
    INDEX
    Explanations

    contrasting or consequential clauses

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

    The neuron flags spans where the writer expresses confusion or asks for clarification—e.g. “I’d never heard of…,” “What does that even mean?,” “What makes you think…?”

    oai_token-act-pair · o4-miniTriggered 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
    踴
    -1.20
    ulating
    -1.13
    autres
    -1.12
     momentous
    -1.10
    aked
    -1.07
     ridiculous
    -1.06
     ridiculously
    -1.05
    ੲ
    -1.05
     ht
    -1.04
     Equals
    -1.04
    POSITIVE LOGITS
     this
    1.55
    Despite
    1.23
    Shortly
    1.13
    ссора
    1.10
    　「
    1.05
    Ironically
    1.05
     massively
    1.04
     pitä
    1.04
     utterly
    1.04
    Since
    1.04
    Activations Density 0.036%

    No Known Activations