© 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. 2666
    Prev
    Next
    INDEX
    Explanations

    phrases or terms associated with misleading information and legal contexts

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

    Mistakes, deception, or false assumptions

    np_acts-logits-general · gemini-2.0-flash

    mistaken belief or deception

    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
    ########.
    -0.52
     assisti
    -0.49
     arrep
    -0.49
    OrEqualTo
    -0.48
    EnableWeb
    -0.46
    hspace
    -0.45
    oncesto
    -0.45
    Kariera
    -0.44
    ByExample
    -0.44
    ällig
    -0.44
    POSITIVE LOGITS
     mistaken
    1.67
     mistakenly
    1.52
     believing
    1.46
     believed
    1.42
     misconception
    1.42
     believe
    1.37
     assumed
    1.37
    以為
    1.35
     misunder
    1.35
     erroneously
    1.34
    Activations Density 0.510%

    No Known Activations