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-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 1612
    Prev
    Next
    INDEX
    Explanations

    affirmative adverbs and phrases indicating certainty or emphasis

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

    Following adverbs and auxiliary verbs

    np_acts-logits-general · gemini-2.0-flash

    certainly, just, only, also followed by verbs

    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
     serem
    -0.55
     charging
    -0.48
     clarification
    -0.46
     okaz
    -0.45
    bit
    -0.44
    .
    -0.44
     consecu
    -0.43
     sighing
    -0.43
     verification
    -0.42
     realisation
    -0.42
    POSITIVE LOGITS
     nahilalakip
    1.15
     can
    1.15
     have
    1.13
     had
    1.10
     cannot
    1.09
     could
    1.09
     will
    1.03
     would
    0.97
     got
    0.97
     has
    0.97
    Activations Density 0.254%

    No Known Activations