© 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. 85196
    Prev
    Next
    INDEX
    Explanations

    desolate landscapes and harsh conditions

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

    The neuron detects names of poets and titles of poems (e.g. “Pound,” “The Waste Land,” “In the Bleak Midwinter,” “Desolation,” etc.).

    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
    ”)
    -1.80
    仍在
    -1.77
    “...
    -1.66
    仍然
    -1.60
     いる
    -1.60
     ajoutez
    -1.59
    只剩下
    -1.58
     impegno
    -1.58
    一旁的
    -1.55
     しま
    -1.54
    POSITIVE LOGITS
    {
    2.09
    ]
    1.55
    气息
    1.49
     Gegend
    1.48
    Semoga
    1.43
    纡
    1.41
     gesucht
    1.38
     I
    1.38
    Therefore
    1.37
     العالمي
    1.36
    Activations Density 0.007%

    No Known Activations