© 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. 125993
    Prev
    Next
    INDEX
    Explanations

    Mess

    np_max-act · gemini-2.0-flash

    This neuron fires on title‐ or header-style text—i.e. section/page titles and headings (like show names with seasons, book or PDF titles, calculator names, schedule headings, etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    references to specific music bands or concerts.

    oai_token-act-pair · gpt-4o-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
     سف
    -0.07
     cupboard
    -0.07
     Tyto
    -0.07
    .hash
    -0.06
    Obsolete
    -0.06
     Surre
    -0.06
     весь
    -0.06
    유
    -0.06
     stabilize
    -0.06
     ти
    -0.06
    POSITIVE LOGITS
     origins
    0.07
    �
    0.06
    etrize
    0.06
    Ark
    0.06
    APIView
    0.06
    timer
    0.06
    !↵↵
    0.06
    สนาม
    0.06
    (PATH
    0.06
     ضر
    0.06
    Activations Density 0.027%

    No Known Activations