© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 8253
    Prev
    Next
    INDEX
    Explanations

    elements related to psychological assessment scales and their reliability

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

    Likert scale anchors and agreement levels

    np_acts-logits-general · gemini-2.0-flash

    describing levels or scales

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     thicken
    -0.59
    prüche
    -0.57
     thicker
    -0.56
    fated
    -0.56
    figsize
    -0.55
    öss
    -0.55
    earlier
    -0.53
    èbre
    -0.53
     thinner
    -0.52
    visation
    -0.52
    POSITIVE LOGITS
     very
    0.78
     completely
    0.76
     extremely
    0.74
    Very
    0.74
     Very
    0.73
     strongly
    0.73
     definitely
    0.71
    very
    0.70
     totally
    0.68
     zcela
    0.67
    Activations Density 0.718%

    No Known Activations