© 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-2-9B
    4. Residual Stream - 16k
    5. 28-GEMMASCOPE-RES-16K
    6. 3871
    Prev
    Next
    INDEX
    Explanations

    frequent references to legal procedures and their implications

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

    Statements of denial or rejection

    np_acts-logits-general · gemini-2.0-flash

    rejection or denial of requests

    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
    worry
    -0.48
     ujednoznacz
    -0.47
     concord
    -0.45
    出版年
    -0.44
     CreateTagHelper
    -0.44
    apunov
    -0.43
     disambiguazione
    -0.43
    createStatement
    -0.43
     يتيمه
    -0.42
     Papi
    -0.42
    POSITIVE LOGITS
     rejection
    1.80
     rejected
    1.66
     reject
    1.64
     rejecting
    1.57
     refusal
    1.57
     refused
    1.51
     Rejection
    1.49
     refus
    1.47
     rejects
    1.45
     refuse
    1.42
    Activations Density 0.854%

    No Known Activations