© 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. 92389
    Prev
    Next
    INDEX
    Explanations

    remain

    np_max-act · gemini-2.0-flash

    This neuron activates on the single verb “remains,” detecting that exact word in its context.

    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
     Katz
    -0.08
    ッシュ
    -0.08
     shoot
    -0.08
     Bishop
    -0.07
     Cisco
    -0.07
     PHOTO
    -0.07
     kinds
    -0.07
    -school
    -0.07
    Ford
    -0.07
    Cisco
    -0.07
    POSITIVE LOGITS
     remains
    0.16
     remain
    0.15
     remained
    0.13
    remain
    0.11
    _remain
    0.10
    amen
    0.08
     remaining
    0.08
    remaining
    0.08
     readdir
    0.08
    .readdir
    0.08
    Activations Density 0.018%

    No Known Activations