© 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
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 28-GEMMASCOPE-RES-16K
    11. 8227
    Prev
    Next
    INDEX
    Explanations

    negative and derogatory descriptors related to people and entities

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

    Negative sentiment and insults

    np_acts-logits-general · gemini-2.0-flash

    rude, vulgar, and aggressive language

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.28.hook_resid_post
    Hook Layer
    28
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     apr
    -0.38
     High
    -0.37
    வை
    -0.37
    UnifiedTopology
    -0.36
    fromJson
    -0.35
     condic
    -0.35
    iläum
    -0.34
     dataSnapshot
    -0.34
    thenReturn
    -0.34
     segu
    -0.34
    POSITIVE LOGITS
     ***!
    0.82
     crap
    0.79
     assholes
    0.77
     idiotic
    0.77
     stupid
    0.76
     dumbass
    0.76
     shit
    0.75
     shite
    0.74
     asshole
    0.73
     idiots
    0.73
    Activations Density 0.259%

    No Known Activations