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 - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 14920
    Prev
    Next
    INDEX
    Explanations

    statements related to security concerns and government actions

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

    Criticism, failings, and inaction

    np_acts-logits-general · gemini-2.0-flash

    delays and inaction

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
     kasarigan
    -0.50
    canestro
    -0.40
    principalColumn
    -0.40
     racc
    -0.38
    клопе
    -0.38
     الحره
    -0.37
     esternos
    -0.37
     sandra
    -0.36
     ddelweddau
    -0.36
    poss
    -0.35
    POSITIVE LOGITS
    webElementXpaths
    0.66
     inaction
    0.51
     delays
    0.45
    cheinend
    0.44
     failings
    0.44
     nedost
    0.44
     insufic
    0.43
     mismanagement
    0.43
     passieren
    0.43
     delay
    0.42
    Activations Density 0.258%

    No Known Activations