Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 3575
    Prev
    Next
    INDEX
    Explanations

    linguistic elements related to movement and direction within a narrative context

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

    Direction or location (into, up, to, inside, home)

    np_acts-logits-general · gemini-2.0-flash

    going into or away from locations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
    AddTagHelper
    -0.72
    enterOuterAlt
    -0.63
    Personendaten
    -0.63
     Audiodateien
    -0.57
    findpost
    -0.57
    KommentareTeilen
    -0.57
    Tembelea
    -0.54
    tagHelperRunner
    -0.53
    AutoField
    -0.52
    EndProject
    -0.52
    POSITIVE LOGITS
    webtoken
    0.37
     yonder
    0.36
     Hiding
    0.36
     hiding
    0.34
     reserved
    0.34
     away
    0.33
     levure
    0.32
     isolado
    0.32
    __(/*!
    0.32
     somewhere
    0.31
    Activations Density 0.245%

    No Known Activations