© 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
    3. Gemma-3-4B-IT
    5. 29-GEMMASCOPE-2-RES-65K
    7. 3999
    Prev
    Next
    INDEX
    Explanations

    Liam Scarlett

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

    instructions specifying JSON output formatting (especially the "Output everything in the following JSON object" phrase and related result-variable/field-format rules).

    oai_token-act-pair · gpt-5-miniTriggered by @chidakshravuru

    tokens related to structured data formatting, field separators, and punctuation that denotes hierarchical organization in complex documents.

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

    technical terms

    np_max-act-logits · claude-4-5-haikuTriggered by @chidakshravuru
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-4b-it/resid_post/layer_29_width_65k_l0_medium
    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
    cantidad
    0.36
    other
    0.34
    additional
    0.33
    Additional
    0.32
    extérieur
    0.32
    also
    0.31
    html
    0.30
    Note
    0.30
    any
    0.30
    note
    0.29
    POSITIVE LOGITS
     chatbots
    0.47
     DevOps
    0.42
     SaaS
    0.40
     chatbot
    0.38
     DirectX
    0.38
     spacetime
    0.38
     ASCII
    0.37
     PCIe
    0.37
     Agile
    0.37
     sentient
    0.37
    Activations Density 0.255%

    No Known Activations