© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 44675
    Prev
    Next
    INDEX
    Explanations

    of

    np_max-act · gemini-2.0-flash

    The neuron detects the assistant’s policy‐style refusal or disclaimer language (e.g. words like “condone,” “promote,” “harmful,” “offensive”) indicating it will not generate disallowed content.

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

    content that promotes or discusses respect and kindness towards individuals of different backgrounds or identities.

    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
     onNext
    -0.06
     Critics
    -0.06
     browsing
    -0.06
    "s
    -0.06
     #####
    -0.06
    	sw
    -0.06
     Bulgarian
    -0.06
     گرف
    -0.06
    load
    -0.06
    MODULE
    -0.06
    POSITIVE LOGITS
    阵
    0.07
     Arbeit
    0.06
    結果
    0.06
     related
    0.06
     MED
    0.06
     Cumhuriyeti
    0.06
     중심
    0.06
    -Tr
    0.06
     دکتر
    0.06
     گروه
    0.06
    Activations Density 0.067%

    No Known Activations