© 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
    7. Residual Stream - 16k
    9. 20-GEMMASCOPE-RES-16K
    11. 12602
    Prev
    Next
    INDEX
    Explanations

    references to dilemmas and challenges related to societal issues and perceptions

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

    expressions of quantity or degree—numbers, percentages, group counts, and intensifiers indicating extent or severity, especially in evaluative or statistical contexts.

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

    qualifying or hedging language that expresses limitations, imperfections, or mild negative assessments.

    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
    httphttps
    -0.48
     <<<<<<<<<<<<<<
    -0.47
    anair
    -0.41
     elm
    -0.40
    Története
    -0.39
     estekak
    -0.39
    enter
    -0.39
     LEM
    -0.39
    AndEndTag
    -0.39
    elm
    -0.38
    POSITIVE LOGITS
    ContentAlignment
    0.49
    ⤹
    0.42
     صوتيه
    0.42
     memand
    0.42
    mejores
    0.40
     ModelExpression
    0.40
    cartão
    0.39
     ktorá
    0.39
    ubscribe
    0.39
     atriz
    0.39
    Activations Density 0.498%

    No Known Activations