© 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-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 1806
    Prev
    Next
    INDEX
    Explanations

    logical contradictions and paradoxes

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

    The neuron strongly activates on dense, technical exposition passages—especially ones full of specialized terminology, formal proofs or comparisons—marking sections where abstract or analytical discussion (e.g. lens‐type comparisons or invariant‐property proofs) dominates.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Also
    0.72
     Thankfully
    0.70
     Luckily
    0.67
     También
    0.66
    อื่นๆ
    0.66
     lainnya
    0.65
     Unfortunately
    0.64
    testimonials
    0.64
     Andere
    0.64
     மற்றும்
    0.64
    POSITIVE LOGITS
     mediating
    0.66
     complementarity
    0.66
     mediated
    0.64
     transcendent
    0.63
     paradigm
    0.63
     radically
    0.62
     interdependence
    0.62
     fundamento
    0.61
     transcendence
    0.61
     abstraction
    0.60
    Activations Density 0.039%

    No Known Activations