Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 10376
    Prev
    Next
    INDEX
    Explanations

    forms of the verb "to do" in various contexts

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

    "I'd" or "would" followed by a verb

    np_acts-logits-general · gemini-2.0-flash

    expressing desires and expectations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
    клопе
    -0.70
    inafter
    -0.57
    pture
    -0.54
    Geplaatst
    -0.52
    EndContext
    -0.52
    queous
    -0.51
    vidla
    -0.49
    Shall
    -0.49
    UAWEI
    -0.49
    matica
    -0.49
    POSITIVE LOGITS
    tagHelperRunner
    0.69
     say
    0.64
     advise
    0.64
    ContentAsync
    0.62
     love
    0.61
     hoped
    0.60
     rather
    0.59
     prefer
    0.58
     øns
    0.57
    SEGUIR
    0.57
    Activations Density 0.072%

    No Known Activations