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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 42323
    Prev
    Next
    INDEX
    Explanations

    text excerpts

    np_max-act · gemini-2.0-flash

    The neuron fires on the “Here” in the stock phrase “Here is my query” (and related prefix tokens) that the user inserts just before stating their actual question.

    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
     pistols
    -0.08
     Jarvis
    -0.08
     всю
    -0.07
    画
    -0.06
     sağlayan
    -0.06
    根
    -0.06
     عص
    -0.06
     postId
    -0.06
    dict
    -0.06
    рел
    -0.06
    POSITIVE LOGITS
    .Speed
    0.06
    TabControl
    0.06
    .setData
    0.06
    mez
    0.06
     sooner
    0.06
     reinforce
    0.06
    _EFFECT
    0.06
     Dhabi
    0.06
     extingu
    0.06
     togg
    0.06
    Activations Density 0.037%

    No Known Activations