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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 104375
    Prev
    Next
    INDEX
    Explanations

    statements and questions related to mathematical evaluations or comparisons

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

    Segment breaks in mathematical question-answer sequences, often appearing between the end of one problem and the beginning of the next one.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
    httphttps
    -1.13
     nakalista
    -0.96
    ConstraintMaker
    -0.93
     تضيفلها
    -0.92
    WriteTagHelper
    -0.88
    AddTagHelper
    -0.85
     ModelExpression
    -0.85
    Personensuche
    -0.84
    transQ
    -0.83
    fromnode
    -0.82
    POSITIVE LOGITS
     Vereine
    0.32
     Zimmermann
    0.32
    ↵↵↵
    0.32
     lost
    0.31
     Cortés
    0.30
    .”
    0.29
     Mak
    0.29
     Shimizu
    0.29
     clubs
    0.29
    Json
    0.29
    Activations Density 0.180%

    No Known Activations