© 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-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 9023
    Prev
    Next
    INDEX
    Explanations

    terms related to formal programs and support initiatives

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

    German language or related to German institutions

    np_acts-logits-general · gemini-2.0-flash

    German language suffixes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ädie
    -0.52
    RegistryLite
    -0.52
     nulle
    -0.47
    Cyfeiriadau
    -0.45
    deutschen
    -0.45
     ones
    -0.43
    ⇓
    -0.42
     SqlCommand
    -0.42
     solchen
    -0.42
    artigen
    -0.42
    POSITIVE LOGITS
    ()]);
    0.66
    "]);
    
    0.63
     bezeichneter
    0.62
    )];
    
    0.60
    oredCriteria
    0.60
    "]];
    0.60
    "]).
    0.60
    Sides
    0.60
    ."));
    0.60
    )}>
    0.59
    Activations Density 0.034%

    No Known Activations