© 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-2-2B
    7. Residual Stream - 16k
    9. 18-GEMMASCOPE-RES-16K
    11. 428
    Prev
    Next
    INDEX
    Explanations

    conditional statements and references to actions or choices

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

    Following certain words in other languages

    np_acts-logits-general · gemini-2.0-flash

    subjunctive verbs / conditional clauses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    !*\
    -0.65
    Geplaatst
    -0.62
     StatefulWidget
    -0.60
    المكان
    -0.59
     خواهید
    -0.57
    ///</
    -0.54
    tometer
    -0.53
     Baillargeon
    -0.53
    وفاته
    -0.52
     trug
    -0.51
    POSITIVE LOGITS
     be
    2.00
    be
    1.30
     soient
    1.19
     Be
    1.16
    Be
    1.15
     puisse
    1.08
     possa
    1.07
     legyen
    1.04
     aient
    1.04
     abbia
    1.03
    Activations Density 0.449%

    No Known Activations