© 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
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 28-GEMMASCOPE-RES-16K
    6. 3359
    Prev
    Next
    INDEX
    Explanations

    expressions of willingness to assist or provide help in various contexts

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

    Following "can," "could," or "will"

    np_acts-logits-general · gemini-2.0-flash

    offering help and services

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.28.hook_resid_post
    Hook Layer
    28
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    şört
    -0.54
     Meksiku
    -0.54
    utives
    -0.47
    LabelTagHelper
    -0.43
    henswürdigkeiten
    -0.42
    CDCl
    -0.42
    rizona
    -0.42
    intios
    -0.42
    iedenis
    -0.41
    transQ
    -0.41
    POSITIVE LOGITS
     help
    0.66
     assist
    0.52
    help
    0.49
    帮你
    0.48
     помочь
    0.46
    幫
    0.46
    帮
    0.45
    幫忙
    0.45
     arrange
    0.43
     discuss
    0.43
    Activations Density 0.249%

    No Known Activations