© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 10266
    Prev
    Next
    INDEX
    Explanations

    instances of the word "interesting" and related expressions that convey curiosity and engagement with the subject matter

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    RTLR
    -0.59
    ConstraintMaker
    -0.57
     noDo
    -0.57
    Ӕ
    -0.54
     Administrativna
    -0.54
    Cyfeiriadau
    -0.50
     Roskov
    -0.50
    Crown
    -0.50
     kasarigan
    -0.50
     gynhyrchwyd
    -0.50
    POSITIVE LOGITS
     interesting
    1.18
    interesting
    1.08
     Interesting
    1.05
    Interesting
    1.05
     intriguing
    1.02
     fascinating
    0.89
     interesante
    0.85
     interessante
    0.82
     interestingly
    0.82
     strange
    0.79
    Activations Density 0.131%

    No Known Activations