© 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. 43432
    Prev
    Next
    INDEX
    Explanations

    chemical synthesis

    np_max-act · gemini-2.0-flash

    This neuron activates on the chemical designation “5-H-ZP” (and its tokenized parts, e.g. “5-H” and “ZP”), i.e. it detects mentions of that specific compound identifier.

    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
    ать
    -0.07
     cheating
    -0.06
    .fx
    -0.06
     όλ
    -0.06
     masturbation
    -0.06
     Decay
    -0.06
     البد
    -0.06
    utomation
    -0.06
    _auto
    -0.05
     Parenthood
    -0.05
    POSITIVE LOGITS
    _items
    0.07
    针
    0.07
     Brazil
    0.07
    ≧
    0.07
    .subtract
    0.07
    ें।
    0.07
    .Node
    0.07
    _radius
    0.07
    ITEM
    0.07
    istribute
    0.07
    Activations Density 0.082%

    No Known Activations