© 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. 18105
    Prev
    Next
    INDEX
    Explanations

    expressions of certainty and affirmation

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

    no followed by other words

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

    the neuron detects strongly opinionated or emphatic words/phrases that signal a speaker's stance or strong sentiment.

    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
    EndInit
    -0.56
    mektedir
    -0.55
    uxxxx
    -0.53
     Meksiku
    -0.48
    rungsseite
    -0.47
    maktadır
    -0.46
     skall
    -0.44
    HasAnnotation
    -0.44
    melidir
    -0.43
    Adapun
    -0.41
    POSITIVE LOGITS
     Aussie
    0.62
     gotta
    0.61
     outta
    0.57
    sient
    0.57
    Gotta
    0.56
    Aussie
    0.56
     auffi
    0.56
     nggak
    0.56
     sorta
    0.56
     somethin
    0.54
    Activations Density 0.383%

    No Known Activations