© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 112490
    Prev
    Next
    INDEX
    Explanations

    code-related syntax elements, such as programming constructs and control flow keywords

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

    computational steps or code blocks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    RTEE
    -0.56
     censiti
    -0.51
     препратки
    -0.51
    quiao
    -0.46
    bital
    -0.46
    itarianism
    -0.45
    fortawesome
    -0.45
    
    -0.45
    ceedings
    -0.44
    featureID
    -0.43
    POSITIVE LOGITS
    いしい
    0.40
    enterOuterAlt
    0.39
    ValueGenerated
    0.35
     Regular
    0.35
    fjspx
    0.34
    ırlı
    0.33
     seduta
    0.33
     recommandée
    0.33
     گیر
    0.33
     оригіналу
    0.32
    Activations Density 0.012%

    No Known Activations