© 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
    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. 6829
    Prev
    Next
    INDEX
    Explanations

    if

    np_max-act · gemini-2.0-flash

    This neuron detects concessive conditional phrases—specifically the “even if it means” construction.

    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
    (bytes
    -0.07
    publish
    -0.07
    you
    -0.07
    ايات
    -0.07
     displaced
    -0.07
    /full
    -0.07
    /fire
    -0.06
     settled
    -0.06
    >:
    -0.06
    .status
    -0.06
    POSITIVE LOGITS
     вв
    0.06
    ?}",
    0.06
     єв
    0.06
    θερ
    0.06
     TF
    0.06
    ертв
    0.06
    whatever
    0.06
     IDF
    0.06
    _CTX
    0.06
     athletes
    0.05
    Activations Density 0.024%

    No Known Activations