Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 86360
    Prev
    Next
    INDEX
    Explanations

    and

    np_max-act · gemini-2.0-flash

    The neuron activates on past‐tense narrative cues—especially auxiliary and main verbs like “was,” “resting,” “greeted,” etc., that signal events already completed.

    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
    Y
    -0.07
    шила
    -0.07
    Africa
    -0.07
    lanır
    -0.07
     ferry
    -0.06
     roman
    -0.06
    Letter
    -0.06
    inn
    -0.06
     hoá
    -0.06
     derecho
    -0.06
    POSITIVE LOGITS
     นาย
    0.07
    _CONSOLE
    0.07
     className
    0.06
     �
    0.06
     h�
    0.06
     dě
    0.06
     alive
    0.06
    คโนโลย
    0.06
     pedestrians
    0.06
    .findViewById
    0.06
    Activations Density 0.116%

    No Known Activations