© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 60889
    Prev
    Next
    INDEX
    Explanations

    Appeals of

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

    The neuron fires on high-level header and label tokens—proper names, place names, docket numbers, property/feature descriptors, court titles, etc.—i.e. key metadata fields in semi-structured document headers.

    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
     California
    -1.46
    California
    -1.41
     Calif
    -1.40
     then
    -1.38
     $-$
    -1.30
    ramento
    -1.27
    Ndr
    -1.23
     theore
    -1.18
    řel
    -1.18
    ตร์
    -1.16
    POSITIVE LOGITS
    																		
    1.52
    																			
    1.34
     "/")
    1.34
    																				
    1.31
     raspberries
    1.26
     handlar
    1.25
    産の
    1.25
     sagt
    1.25
    ########
    1.23
    																
    1.23
    Activations Density 0.400%

    No Known Activations