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-2-9B
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 70671
    Prev
    Next
    INDEX
    Explanations

    sentences expressing knowledge or realization

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

    Form of "know" or similar cognition verbs indicating internal certainty about a situation, often followed by content describing what the person knows or understands with confidence.

    eleuther_acts_top20 · claude-3-7-sonnet-20250219Triggered by @lu-christina
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
     Facilities
    -0.47
     facilities
    -0.47
    Autoritní
    -0.45
    égard
    -0.45
     leçons
    -0.45
     Speise
    -0.45
     incidente
    -0.44
     parut
    -0.44
     unterkunft
    -0.43
     borboleta
    -0.43
    POSITIVE LOGITS
     know
    0.62
    know
    0.56
    withIdentifier
    0.55
     recognise
    0.54
    我知道
    0.53
     KNOW
    0.52
     recognize
    0.51
     knows
    0.49
    知道
    0.47
    knows
    0.47
    Activations Density 0.034%

    No Known Activations