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 - 16k
    5. 8-GEMMASCOPE-RES-16K
    6. 134
    Prev
    Next
    INDEX
    Explanations

    phrases or sentences that express or imply ongoing actions, states, or situations. It seems to activate strongly for phrases indicating something that is currently happening, will happen, or is likely to happen, often using present tense verbs or modal verbs like "will", "is to", or "are likely to".

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @danielepoterti

    statements related to personal experiences and feelings of struggle

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 8-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_8/width_16k/average_l0_99
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.8.hook_resid_post
    Hook Layer
    8
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     was
    -0.52
     meant
    -0.36
     belonged
    -0.35
     avait
    -0.35
     had
    -0.35
    もありました
    -0.34
     vardı
    -0.34
     consisted
    -0.33
     took
    -0.33
     did
    -0.33
    POSITIVE LOGITS
     الحره
    0.69
    featureID
    0.66
    Климат
    0.66
    دانشنامهٔ
    0.64
    SourceChecksum
    0.63
    +#+#
    0.61
    хьтан
    0.59
     بيها
    0.59
     attualmente
    0.59
     nyní
    0.58
    Activations Density 2.753%

    No Known Activations