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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & 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. 1121
    Prev
    Next
    INDEX
    Explanations

    events, dates, and times related to meetings or presentations

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

    Times of day, often followed by "at" or "from"

    np_acts-logits-general · gemini-2.0-flash

    time of day

    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
    featureID
    -0.63
    esez
    -0.58
     GetEnumerator
    -0.53
     незавершена
    -0.53
    цездатний
    -0.52
     تضيفلها
    -0.51
     cref
    -0.51
    Jereo
    -0.51
     insuffisamment
    -0.50
     iconTwitter
    -0.50
    POSITIVE LOGITS
     pukul
    1.01
     klokken
    0.90
     ساعت
    0.88
     noon
    0.87
     starting
    0.86
     early
    0.84
     around
    0.84
     beginning
    0.80
     klock
    0.79
     Noon
    0.77
    Activations Density 0.220%

    No Known Activations