© 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. 5119
    Prev
    Next
    INDEX
    Explanations

    error handling statements and return conditions in code

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

    programming error handling and error checking code (specifically in Go/C-like languages).

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

    code patterns that signal error-handling paths, such as checking for failures and then reporting, returning, or jumping to cleanup.

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

    error handling code patterns, particularly where errors are checked and control flow is redirected (return statements, goto statements, or error propagation).

    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
     FetchType
    -0.66
     betweenstory
    -0.64
     مشين
    -0.61
    UserScript
    -0.61
     CreateTagHelper
    -0.60
     fieldNum
    -0.59
     CURIAM
    -0.59
    
    -0.58
    RotationOrder
    -0.58
    styleType
    -0.58
    POSITIVE LOGITS
     powod
    0.35
     Spuren
    0.33
     kehilangan
    0.32
    <code>
    0.31
     insbesondere
    0.29
     Genehmigung
    0.29
     виправи
    0.28
    그러나
    0.28
     could
    0.28
    principalTable
    0.28
    Activations Density 0.068%

    No Known Activations