© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 9-GEMMASCOPE-RES-131K
    11. 94693
    Prev
    Next
    INDEX
    Explanations

    instances of specific keywords and phrases related to scientific or technical contexts

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

    such as, anti-PD-1, well-known

    np_acts-logits-general · gemini-2.5-flash-lite

    the neuron mainly detects numeric tokens (numbers, numerals, and numeric/decimal tokens used for years, percentages, measurements, or references).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.9.hook_resid_post
    Hook Layer
    9
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     CreateTagHelper
    -1.08
    AddTagHelper
    -0.93
     ***!
    -0.85
    GOTREF
    -0.80
     autorytatywna
    -0.77
    tagHelperRunner
    -0.77
     виправивши
    -0.77
     zwiſchen
    -0.75
    ஦
    -0.73
    [@BOS@]
    -0.73
    POSITIVE LOGITS
    class
    0.28
    include
    0.27
     Europe
    0.27
     vzor
    0.26
     moky
    0.25
    u
    0.25
     laik
    0.25
    janjian
    0.25
    d
    0.25
    ...
    0.25
    Activations Density 0.135%

    No Known Activations