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. 20-GEMMASCOPE-RES-131K
    6. 73553
    Prev
    Next
    INDEX
    Explanations

    programming or technical references and annotations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     colours
    -0.46
    hiculo
    -0.43
    Colours
    -0.43
     bull
    -0.42
    colours
    -0.41
     coloured
    -0.40
    geslacht
    -0.39
     ugly
    -0.38
    つぶ
    -0.38
    що
    -0.38
    POSITIVE LOGITS
    Autowired
    1.02
    StoryboardSegue
    0.89
    CppMethod
    0.82
    MockBean
    0.74
    ArgsConstructor
    0.71
    InjectMocks
    0.64
    BindView
    0.62
    GTCX
    0.62
    pushFollow
    0.58
    ThroughAttribute
    0.57
    Activations Density 0.002%

    No Known Activations