© 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. 55798
    Prev
    Next
    INDEX
    Explanations

    code related text

    np_max-act · gemini-2.0-flash

    The neuron responds to relative‐clause markers—especially the words “that” and “which”—used to introduce explanatory or subordinate clauses.

    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
    'clock
    -0.08
    (x
    -0.07
     theres
    -0.07
     WI
    -0.06
    (lhs
    -0.06
    окс
    -0.06
    сутств
    -0.06
     Rem
    -0.06
    _CTX
    -0.06
    (seq
    -0.06
    POSITIVE LOGITS
     emphas
    0.07
    ัวร
    0.07
    opro
    0.06
    /connection
    0.06
     gdzie
    0.06
     söy
    0.06
     Dě
    0.06
     mysql
    0.06
     đúng
    0.06
    imension
    0.06
    Activations Density 0.179%

    No Known Activations