© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 73789
    Prev
    Next
    INDEX
    Explanations

    Questions and answers

    np_max-act · gemini-2.0-flash

    sentences that express opinions or sentiments.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron strongly activates on the model’s internal control markers (e.g. `<|start_header_id|>`, `<|end_header_id|>`, `<|eot_id|>`), i.e. structural/meta tokens rather than the user’s words.

    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
     pra
    -0.07
     notifications
    -0.07
     indx
    -0.07
     Recording
    -0.07
    EdgeInsets
    -0.06
     tab
    -0.06
     Pla
    -0.06
     Levi
    -0.06
     keys
    -0.06
     SPD
    -0.06
    POSITIVE LOGITS
    люд
    0.06
     탈
    0.06
    γά
    0.06
    طح
    0.06
    .EX
    0.06
    sonian
    0.06
     �
    0.06
     ($.
    0.06
    oxic
    0.06
    Scheduler
    0.06
    Activations Density 0.024%

    No Known Activations