© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 6286
    Prev
    Next
    INDEX
    Explanations

    truth

    np_max-act · gemini-2.0-flash

    This neuron detects discourse markers or adverbial intensifiers (words like “actually,” “really,” “indeed,” “in reality,” etc.) that signal emphasis, correction, or contrast.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    }},
    -0.07
    ันอ
    -0.06
    TypeDef
    -0.06
     mutlaka
    -0.06
     officials
    -0.06
    confirmed
    -0.06
     FC
    -0.06
     Prospect
    -0.06
     ا
    -0.06
     Buccaneers
    -0.06
    POSITIVE LOGITS
     самом
    0.07
    .sun
    0.07
    .tail
    0.07
    紧
    0.06
    .instances
    0.06
     initialization
    0.06
     SELF
    0.06
    fu
    0.06
    ActiveSheet
    0.06
     controlling
    0.06
    Activations Density 0.023%

    No Known Activations