© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 31-GEMMASCOPE-RES-16K
    11. 3043
    Prev
    Next
    INDEX
    Explanations

    references to specific types of scientific or technical information, particularly related to gray-scale concepts and supplemental materials in various contexts

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot

    proper nouns, particularly names of organizations, institutions, or technical terms in academic or legal documents.

    oai_token-act-pair · claude-3-7-sonnet-20250219
    Triggered by @neilrathi

    proper nouns and capitalized words, particularly names of specific entities, institutions, and places.

    oai_token-act-pair · claude-4-5-haiku
    Triggered by @emiglarou
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.31.hook_resid_post
    Hook Layer
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Negative Logits
     fertiliser
    -0.57
    KommentareTeilen
    -0.57
     nahilalakip
    -0.55
    paedia
    -0.53
    ്
    -0.52
     diarrhoea
    -0.52
    Fibre
    -0.52
     optimised
    -0.52
    Personendaten
    -0.51
     honoured
    -0.51
    POSITIVE LOGITS
    ency
    0.56
     StringTokenizer
    0.49
    ers
    0.48
    es
    0.48
     poveznice
    0.48
     confiable
    0.46
    an
    0.45
    logic
    0.45
    vezető
    0.44
     Pacífico
    0.44
    Activations Density 2.414%

    No Known Activations