Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 9-GEMMASCOPE-RES-131K
    6. 58325
    Prev
    Next
    INDEX
    Explanations

    mentions of tricks and hacks

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

    clever methods and tricks

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     StatefulWidget
    -0.57
     تانيه
    -0.49
     شع
    -0.48
     UNF
    -0.47
     Affirm
    -0.47
    CAF
    -0.47
    afd
    -0.46
     Autorizaciones
    -0.46
    ConstraintMaker
    -0.46
     Confer
    -0.45
    POSITIVE LOGITS
     tricks
    1.85
     Tricks
    1.71
    tricks
    1.48
     trucos
    1.39
     astuces
    1.02
     trick
    0.88
     techniques
    0.78
     truco
    0.77
     TECHNIQUES
    0.76
     tactics
    0.75
    Activations Density 0.002%

    No Known Activations