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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. DeepSeek-R1-Dist-Llama-8B
    3. 15-LLAMASCOPE-SLIMPJ-RES-32K
    4. 4460
    Prev
    Next
    INDEX
    Explanations

    phrases related to public relations and contractual agreements

    oai_token-act-pair · gpt-4o-mini
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    fnlp/Llama-Scope-R1-Distill/800M-Slimpajama-0-OpenR1-Math-220k/L15R
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Hzfinfdu/SlimPajama-3B
    Features
    32,768
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    sector
    -0.15
    boss
    -0.14
    PÅĻed
    -0.13
    PÅĻi
    -0.13
    غÙĨ
    -0.13
    onomy
    -0.13
    çªģ
    -0.13
    åī¯
    -0.13
    pig
    -0.13
    oids
    -0.13
    POSITIVE LOGITS
     firm
    0.19
     hired
    0.17
     retained
    0.17
     external
    0.15
     firma
    0.15
    arges
    0.15
    edis
    0.15
     Firm
    0.14
     firms
    0.14
    neys
    0.14
    Activations Density 0.054%

    No Known Activations