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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 131k
    5. 24-GEMMASCOPE-RES-131K
    6. 1083
    Prev
    Next
    INDEX
    Explanations

    phrases related to theft and missing items

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

    taking, stolen, removed, unauthorized access

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_24/width_131k/average_l0_97
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    enderror
    -0.56
     հղումներ
    -0.54
    enterOuterAlt
    -0.51
     ролях
    -0.51
    ReusableCell
    -0.50
    cinnati
    -0.50
    endpush
    -0.47
    autoreleasepool
    -0.46
    OPSIS
    -0.46
    addCriterion
    -0.46
    POSITIVE LOGITS
     stealing
    0.60
     steal
    0.57
     steals
    0.50
     theft
    0.49
     thefts
    0.46
     ब्रेकडाउन
    0.45
     thief
    0.45
     unauthorized
    0.42
    steal
    0.42
     sneaking
    0.41
    Activations Density 0.030%

    No Known Activations