© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 8956
    Prev
    Next
    INDEX
    Explanations

    intoxication

    np_max-act · gemini-2.0-flash

    The neuron activates on Italian word fragments referring to smoking (e.g. “fumare,” “fumo,” “canne”), so it’s detecting mentions of smoking.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    Finds mentions of intoxication or substance use (being high/drunk, smoking, or related drug/alcohol contexts).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    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
    otence
    -0.07
    委员会
    -0.07
     коллек
    -0.06
    ifes
    -0.06
     yell
    -0.06
     MON
    -0.06
     Polic
    -0.06
    _eff
    -0.06
    _worker
    -0.06
    Injection
    -0.06
    POSITIVE LOGITS
     –↵
    0.06
     provoc
    0.06
     queryInterface
    0.06
     každý
    0.06
    -SA
    0.06
     microsoft
    0.06
     fflush
    0.06
    .link
    0.06
     평가
    0.06
    .goto
    0.06
    Activations Density 0.036%

    No Known Activations