© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-27B-IT
    3. 3-GEMMASCOPE-2-TRANSCODER-262K
    4. 38042
    Prev
    Next
    INDEX
    Explanations

    **TOP_ACTIVATING_TEXTS**: * "The Rebel Alliance:" (followed by descriptions of fighting for freedom, rebellion) * "Rebel Alliance fighting for freedom" * "Rebel Alliance (later)" * "Rebel Alliance." (plans for Empire's weapon) * "Republic Square:" (description of restaurants/cafes) * "Marcelo Rebelo de Sousa." (President of Portugal) * "Republic of Singapore" (country name) * "Singapore" (city-state) * "rebelalliance" (mentioned in a security score context) * "Serb Position (primarily Republika Srpska):" (political context)Let's analyze the patterns:* **Alliance/alliance**: Appears very frequently in `MAX_ACTIVATING_TOKENS` and `TOP_ACTIVATING_TEXTS`. It's often associated with "Rebel Alliance" and themes of fighting for freedom.* **Square**: Appears in `MAX_ACTIVATING_TOKENS`. In `TOP_ACTIVATING_TEXTS`, "Republic Square" is mentioned as a location for restaurants.* **Sousa**: Appears in `MAX_ACTIVATING_TOKENS`. In `TOP_ACTIVATING_TEXTS`, "Marcelo Rebelo de Sousa" is identified as the President of Portugal.* **Singapore**: Appears frequently in `MAX_ACTIVATING_TOKENS`.*Rebel Alliance*, Sousa, Singapore, Srps, Square

    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_3_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
    an
    0.83
    ان
    0.71
    ח
    0.70
    ק
    0.68
    as
    0.68
     sensitized
    0.67
    다
    0.67
    १
    0.64
    க
    0.62
    ER
    0.57
    POSITIVE LOGITS
    </h3>
    0.65
    یت
    0.61
    mere
    0.61
     Estates
    0.59
    বে
    0.58
     Capitals
    0.58
    বিদ্য
    0.58
    бить
    0.57
    itis
    0.57
    mati
    0.57
    Activations Density 0.000%

    No Known Activations