© 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-2-9B-IT
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 464
    Prev
    Next
    INDEX
    Explanations

    instances of the verb "was."

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

    Preceding "searching", "trying", "able", "also", etc. after the word "was"

    np_acts-logits-general · gemini-2.0-flash

    past tense auxiliary "was" or "were"

    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-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 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
    <bos>
    -0.56
     ſtate
    -0.46
     houſe
    -0.44
     ainfi
    -0.43
    iffion
    -0.42
    outheast
    -0.42
     ſtre
    -0.41
     démoc
    -0.39
    Rispondi
    -0.39
    southeast
    -0.39
    POSITIVE LOGITS
     was
    1.09
    was
    0.98
     were
    0.98
     Was
    0.93
    Was
    0.90
    were
    0.88
     WERE
    0.81
     Were
    0.80
     было
    0.79
     WAS
    0.79
    Activations Density 0.485%

    No Known Activations