© 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. 6697
    Prev
    Next
    INDEX
    Explanations

    terms related to technology conferences and the field of audio forensics

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

    technical and professional conference details, especially related to technology, software development, and academic events.

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

    dates and numerical information in technical or conference contexts.

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

    descriptions of technical or academic conferences and professional events, especially in computing, engineering, and related research domains.

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

    professional and technical content related to conferences, research, technology, and academic or industry events.

    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
     purpoſe
    -0.53
    GTCX
    -0.49
     fédé
    -0.44
     myſelf
    -0.44
     itſelf
    -0.44
     ſtate
    -0.43
     couvercle
    -0.41
     preſent
    -0.40
     <<<<<<<<<<<<<<
    -0.40
     oprot
    -0.40
    POSITIVE LOGITS
    WebElementEntity
    0.57
    AddTagHelper
    0.52
     practition
    0.48
     attendees
    0.48
    ніципалі
    0.47
     topics
    0.46
     careers
    0.46
     altern
    0.46
     accounting
    0.45
     accountants
    0.45
    Activations Density 0.140%

    No Known Activations