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. 40547
    Prev
    Next
    INDEX
    Explanations

    old

    np_max-act · gemini-2.0-flash

    narrative elements of fantasy adventures.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron activates on fantasy‐flavored descriptive words and fragments (e.g. “mágico,” “antiguas,” “legendarias,” “oculto,” “ritual,” etc.), i.e. it’s detecting magical or mystical story elements.

    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
     TNT
    -0.07
     endurance
    -0.06
     visibility
    -0.06
     rif
    -0.06
     gum
    -0.06
    ับน
    -0.06
     NaN
    -0.06
     bondage
    -0.06
    ceeded
    -0.06
     też
    -0.06
    POSITIVE LOGITS
    ových
    0.07
    слід
    0.06
     Pose
    0.06
    APON
    0.06
    _SPELL
    0.06
    リー
    0.06
    elia
    0.06
     Crimes
    0.06
     Комп
    0.06
    ักษณ
    0.06
    Activations Density 0.037%

    No Known Activations