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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 6358
    Prev
    Next
    INDEX
    Explanations

    instances of the word "cut" and its derivatives indicating actions of reduction or severing

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

    Sentences containing forms of the word "cut"

    np_acts-logits-general · gemini-2.0-flash

    cut off, cut out, cut coverage

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    atoare
    -0.47
     headquarters
    -0.43
     being
    -0.41
     who
    -0.41
     faptul
    -0.41
     valueForKey
    -0.41
     servici
    -0.40
    等级
    -0.40
     szint
    -0.40
     Schaefer
    -0.39
    POSITIVE LOGITS
    Cutting
    1.23
     Cutting
    1.19
     cut
    1.15
     CUT
    1.14
    Cut
    1.13
    cut
    1.13
    cutting
    1.12
     Cut
    1.11
     cutting
    1.09
    cuts
    1.09
    Activations Density 0.160%

    No Known Activations