© 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 - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 7687
    Prev
    Next
    INDEX
    Explanations

    phrases related to connections and interactions

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

    causal relationships indicated by words like "because," "could," "may," and "due to."

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    language that expresses causal or consequential relationships with modal or qualifying clauses indicating possibility and outcomes.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    phrases indicating potential problems, failures, or negative consequences that could occur.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    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
    MessageOf
    -0.70
    ✭✭
    -0.57
    DotNetBar
    -0.55
    oredCriteria
    -0.53
     bezeichneter
    -0.53
     betweenstory
    -0.51
    oneofs
    -0.49
    BufferException
    -0.48
    Datuak
    -0.48
    ConstraintMaker
    -0.48
    POSITIVE LOGITS
     Económica
    0.40
     représ
    0.38
     contribue
    0.37
     coraz
    0.36
    Contribu
    0.36
     mahdol
    0.36
     inmobili
    0.35
     ralla
    0.34
     prefeitura
    0.34
     bronce
    0.33
    Activations Density 0.449%

    No Known Activations