© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 9-GEMMASCOPE-RES-131K
    11. 38514
    Prev
    Next
    INDEX
    Explanations

    instructions or guidelines

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    how to verbs

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

    The neuron detects the word "to" when used as the infinitival/marker in "How to" or "to + verb" constructions.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.9.hook_resid_post
    Hook Layer
    9
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    EndContext
    -0.69
     Мексичка
    -0.66
    isContained
    -0.59
    
    -0.59
    jspb
    -0.57
    pexpr
    -0.55
     CURIAM
    -0.54
     Infórmanos
    -0.52
    UnknownFieldSet
    -0.49
    WriteLiteral
    -0.49
    POSITIVE LOGITS
    
    0.52
     effective
    0.44
    ljiv
    0.42
     krach
    0.40
    UrlResolution
    0.39
     Dot
    0.38
     EFFECTIVE
    0.38
     epä
    0.36
     przewod
    0.36
    effective
    0.36
    Activations Density 0.177%

    No Known Activations