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. 20-GEMMASCOPE-RES-131K
    6. 8401
    Prev
    Next
    INDEX
    Explanations

    verbs related to research and investigation

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

    reporting research findings

    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
     ſch
    -0.93
    ſelf
    -0.88
     myſelf
    -0.86
     ſta
    -0.82
    ſelves
    -0.79
     ſte
    -0.77
     Beſ
    -0.76
    fjspx
    -0.75
     Perſ
    -0.75
     ſche
    -0.74
    POSITIVE LOGITS
     and
    0.42
     by
    0.40
     in
    0.37
     while
    0.36
    .
    0.35
     vlastnosti
    0.34
     relative
    0.34
     via
    0.34
     clairement
    0.33
     as
    0.32
    Activations Density 0.098%

    No Known Activations