Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Gemma-3-4B-IT
    3. 25-GEMMASCOPE-2-TRANSCODER-262K
    4. 8576
    Prev
    Next
    INDEX
    Explanations

    force

    np_max-act-logits · gemini-2.5-flash-lite

    Based on the activation patterns across all the text samples, this neuron activates strongly on **first-person narrative perspective and introspective emotional states**, particularly when characters are processing complex feelings, memories, or moments of vulnerability. The neuron shows high activations on pronouns like "I

    oai_token-act-pair · claude-4-5-haikuTriggered by @vahramatayan

    narrative text indicating first-person perspective or character actions, particularly in role-play, dialogue, or story contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @vahramatayan
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-4b-it/transcoder_all/layer_25_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
    Conteudos
    0.75
     Toward
    0.74
    antam
    0.68
    Toward
    0.68
     Towards
    0.67
     Townsend
    0.67
    ainted
    0.66
    Horn
    0.65
     וב
    0.65
     towards
    0.65
    POSITIVE LOGITS
     forcing
    0.83
     force
    0.79
     apre
    0.76
     fingerprint
    0.76
     FORCE
    0.74
     resett
    0.71
    forcing
    0.71
     base
    0.70
     শ্বশুর
    0.70
     Force
    0.68
    Activations Density 0.195%

    No Known Activations