© 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. 34-GEMMASCOPE-RES-131K
    4. 69910
    Prev
    Next
    INDEX
    Explanations

    idea and description of what

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

    The neuron is looking for numeric tokens—especially multi-digit numbers (e.g. “1024,” “2128,” “1752,” etc.).

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

    abstract, expository discourse markers that frame or describe information (concepts, perspectives, summaries) rather than concrete content.

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

    words or phrases that introduce examples, illustrations, or detailed explanations.

    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_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
    ⎡
    -0.82
     Seward
    -0.78
     Foxes
    -0.76
     Ș
    -0.76
     Thir
    -0.75
    cony
    -0.74
    ABAJO
    -0.73
     nomes
    -0.72
     tristesse
    -0.72
    																										
    -0.71
    POSITIVE LOGITS
     of
    1.87
     مما
    1.18
     لما
    1.15
     того
    0.88
     siitä
    0.86
     فيما
    0.85
     במה
    0.84
     بما
    0.84
     about
    0.83
     několika
    0.81
    Activations Density 0.070%

    No Known Activations