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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 18-GEMMASCOPE-RES-16K
    6. 14611
    Prev
    Next
    INDEX
    Explanations

    modal verbs and expressions of desire or preference

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

    Following pronouns with what they want/like/see

    np_acts-logits-general · gemini-2.0-flash

    pronoun + verb of desire/liking

    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
     Zep
    -0.53
    AFX
    -0.47
     emin
    -0.46
    Inc
    -0.44
    roffen
    -0.44
    ogos
    -0.43
    이는
    -0.42
    inc
    -0.42
     Inc
    -0.41
    ండ
    -0.41
    POSITIVE LOGITS
     desired
    1.26
    desired
    1.10
     desire
    1.08
     wished
    1.01
    desire
    0.98
     wish
    0.95
     Desired
    0.93
     desires
    0.91
     wishes
    0.89
     quiser
    0.89
    Activations Density 0.264%

    No Known Activations