© 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. 16-GEMMASCOPE-RES-16K
    11. 1834
    Prev
    Next
    INDEX
    Explanations

    phrases that introduce contrast or exceptions

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

    Follows "but" indicating a contrast

    np_acts-logits-general · gemini-2.0-flash

    not ... but clauses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Roskov
    -0.79
     насељу
    -0.72
    Переваги
    -0.71
     Pliny
    -0.61
     AIN
    -0.59
     eto
    -0.58
     fisuras
    -0.58
     ElementRef
    -0.58
     Laine
    -0.57
    fromnode
    -0.57
    POSITIVE LOGITS
     indeed
    0.81
     sondern
    0.73
     vielmehr
    0.71
    indeed
    0.66
     certainly
    0.65
     perhaps
    0.64
    むしろ
    0.64
     anzi
    0.63
     melainkan
    0.61
     simply
    0.60
    Activations Density 0.099%

    No Known Activations