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. 3780
    Prev
    Next
    INDEX
    Explanations

    expressions of anticipation and excitement about upcoming events or experiences

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

    Expressing anticipation or looking ahead

    np_acts-logits-general · gemini-2.0-flash

    looking forward to events

    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
    archiviato
    -0.58
    trand
    -0.56
    FirstResponder
    -0.55
    itzender
    -0.55
     parietal
    -0.51
     Instantiate
    -0.51
    iStock
    -0.51
    jach
    -0.51
    fwrite
    -0.49
    }],
    
    -0.48
    POSITIVE LOGITS
     anticipation
    2.07
     eagerly
    1.74
     anticipating
    1.57
     anticip
    1.56
     excitement
    1.53
     anticipate
    1.44
     anxiously
    1.43
     anticipated
    1.42
     awaited
    1.41
     excited
    1.40
    Activations Density 0.231%

    No Known Activations