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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 5941
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and symbols commonly used in equations or statistical notation

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

    mathematical notation and symbols from academic/scientific text.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    mathematical notation variables and symbols within equations, particularly focusing on variables like x, y, and t.

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

    mathematical LaTeX-style equations and notation, especially calculus/probability expressions with integrals, differentials, and variables using subscripts or superscripts.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    mathematical variables and symbols in LaTeX equations.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    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
    ReusableCell
    -1.02
    WriteTagHelper
    -0.96
    posedge
    -0.96
    TagMode
    -0.96
     CreateTagHelper
    -0.89
    :✨
    -0.88
    ########.
    -0.88
    RenderAtEndOf
    -0.88
    ParallelGroup
    -0.84
     surla
    -0.84
    POSITIVE LOGITS
     is
    0.44
    ,
    0.40
     will
    0.34
     while
    0.33
     as
    0.33
    I
    0.31
     conmigo
    0.31
    and
    0.31
     and
    0.31
     because
    0.31
    Activations Density 1.851%

    No Known Activations