© 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. 89916
    Prev
    Next
    INDEX
    Explanations

    account and order information

    np_max-act · gemini-2.0-flash

    The neuron fires on placeholder tokens and key identifiers—especially customer or agent name tokens (e.g. NAME_n) and the word “transaction” (and its associated number)—i.e. it spots name‐ and transaction‐related fields.

    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
     metic
    -0.07
     Irene
    -0.07
    ículo
    -0.07
    MK
    -0.06
     деятель
    -0.06
     DISPLAY
    -0.06
     будто
    -0.06
    漢
    -0.06
    	RE
    -0.06
     gamm
    -0.06
    POSITIVE LOGITS
     pregnancy
    0.07
     또는
    0.07
    _SELECTED
    0.07
    <const
    0.07
     Lexer
    0.06
    ('__
    0.06
    .Promise
    0.06
    (document
    0.06
    (song
    0.06
    咨
    0.06
    Activations Density 0.001%

    No Known Activations