Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 54700
    Prev
    Next
    INDEX
    Explanations

    quantum physics

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on content-bearing, multi-syllable scientific or technical terms (nouns, adjectives, and verbs) characteristic of research-paper prose.

    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
    prefs
    -0.07
    rne
    -0.06
     Pierce
    -0.06
    IDX
    -0.06
     Model
    -0.06
     Mara
    -0.06
    	ax
    -0.06
     sahiptir
    -0.06
    ceptors
    -0.06
    TOCOL
    -0.06
    POSITIVE LOGITS
    ",'
    0.06
    λιο
    0.06
    歯
    0.06
    ”?
    0.06
     позвол
    0.06
    くれた
    0.06
    من
    0.06
    ؟
    0.06
    inner
    0.06
     adapted
    0.06
    Activations Density 0.018%

    No Known Activations