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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 10362
    Prev
    Next
    INDEX
    Explanations

    proper nouns and specific locations

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

    references to formal news-style named entities—especially UK institutions, locations, dates/times, and court-related attributions—often accompanied by colons and other reporting punctuation.

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

    proper nouns and formal titles in news articles or official documents, particularly location names, institutional names, and quoted speech.

    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
     pot
    -0.42
     ver
    -0.40
     Thomson
    -0.35
     ange
    -0.34
     techo
    -0.33
    しの
    -0.33
     Fer
    -0.32
    indi
    -0.32
     particular
    -0.32
     hol
    -0.32
    POSITIVE LOGITS
    Personendaten
    0.80
    UnusedPrivate
    0.72
     HasFactory
    0.72
    expandindo
    0.72
    tagHelperRunner
    0.71
    IBOutlet
    0.70
     propOrder
    0.68
     desmotivaciones
    0.68
    fjspx
    0.68
     Geſch
    0.66
    Activations Density 0.021%

    No Known Activations