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. 41866
    Prev
    Next
    INDEX
    Explanations

    continue

    np_max-act · gemini-2.0-flash

    chemical companies with a focus on sustainability and environmental responsibility.

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

    This neuron activates on the verb “continue” (especially the past‐tense form “continued”) indicating mentions of ongoing or resumed actions.

    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
     sf
    -0.06
     axes
    -0.06
    189
    -0.06
     mach
    -0.06
     Baxter
    -0.06
     respondents
    -0.06
     jumps
    -0.06
     lamps
    -0.06
    Distribution
    -0.06
     sheds
    -0.06
    POSITIVE LOGITS
     continue
    0.12
     continued
    0.12
     continuing
    0.11
     continues
    0.11
    continued
    0.10
     Continuing
    0.09
    went
    0.09
     продовж
    0.09
    .chk
    0.08
     продолж
    0.08
    Activations Density 0.030%

    No Known Activations