© 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. 43046
    Prev
    Next
    INDEX
    Explanations

    harmful chemicals

    np_max-act · gemini-2.0-flash

    The neuron fires on words and phrases that signal chemical hazards or toxicity concerns (e.g. “toxic,” “harmful,” “proven,” “carcinogen,” “concern”).

    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
    θα
    -0.08
    lio
    -0.07
     Ryan
    -0.06
     SME
    -0.06
    ぬ
    -0.06
    yan
    -0.06
     Janet
    -0.06
     연락
    -0.06
    Time
    -0.06
     Chad
    -0.06
    POSITIVE LOGITS
    international
    0.07
     END
    0.07
     vešker
    0.06
    %',
    0.06
    ?“
    0.06
    。」
    0.06
    firebase
    0.06
     CONT
    0.06
    ertools
    0.06
    (Parse
    0.06
    Activations Density 0.036%

    No Known Activations