© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 129691
    Prev
    Next
    INDEX
    Explanations

    many

    np_max-act · gemini-2.0-flash

    This neuron fires on clause- or sentence-initial discourse markers and adverbial connectors (e.g. “Conveying,” “Just,” “Because,” “so,” “After,” “trying,” etc.) that introduce explanations or transitions.

    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
     Hipp
    -0.07
     startups
    -0.07
    чих
    -0.07
     PŘ
    -0.07
     신청
    -0.07
     Gate
    -0.06
    ляться
    -0.06
    不安
    -0.06
    犬
    -0.06
    .Subscribe
    -0.06
    POSITIVE LOGITS
    //$
    0.06
    }?
    0.06
    ”?
    0.06
    $headers
    0.06
     manifesto
    0.06
     νεφοκάλυψης
    0.06
     =$
    0.06
     CONTRACT
    0.06
    atica
    0.06
    annual
    0.06
    Activations Density 0.334%

    No Known Activations