© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-27B-IT
    3. 24-GEMMASCOPE-2-TRANSCODER-262K
    4. 175204
    Prev
    Next
    INDEX
    Explanations

    inferred**Analyzing Neuron Behavior**The `MAX_ACTIVATING_TOKENS` seem to be common words or punctuation that can appear anywhere.The `TOKENS_AFTER_MAX_ACTIVATING_TOKEN` provides more context. Let's see if there's a pattern between them.1.**Analyzing Neuron Behavior**I'm looking for a concise explanation (3-20 words) of what this neuron detects, focusing on patterns in the provided lists.* **MAX_ACTIVATING_TOKENS**: `was`, `was

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/transcoder_all/layer_24_width_262k_l0_small_affine
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    し
    0.52
    ια
    0.44
     humains
    0.43
    疃
    0.42
    rapa
    0.42
    联赛
    0.41
     alır
    0.41
    ંધ
    0.41
    க்கெ
    0.41
    یره
    0.41
    POSITIVE LOGITS
    ing
    0.54
     l
    0.50
     climb
    0.50
     предусматри
    0.49
     прогноз
    0.48
     habil
    0.47
     conviction
    0.46
    reg
    0.45
    ff
    0.44
    Accounts
    0.44
    Activations Density 0.000%

    No Known Activations