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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 1876
    Prev
    Next
    INDEX
    Explanations

    different verb forms, particularly those in the present and past continuous tense

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

    verb endings indicating ongoing action or state

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    })*/
    -0.65
    \}\\
    -0.55
     muligt
    -0.51
    assertThrows
    -0.51
     cholera
    -0.50
    Soda
    -0.49
    */
    
    
    -0.49
     Ulm
    -0.49
     Blitz
    -0.49
     Kitz
    -0.49
    POSITIVE LOGITS
    めている
    0.63
     てい
    0.63
    している
    0.63
    っている
    0.63
    れている
    0.60
    を持っている
    0.59
    まっている
    0.56
     bezig
    0.55
    えている
    0.55
    をしている
    0.54
    Activations Density 0.007%

    No Known Activations