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. 11-LLAMASCOPE-SLIMPJ-OPENR1-RES-32K
    4. 8420
    Prev
    Next
    INDEX
    Explanations

    chunks of text generated by large language models when asked to think step by step

    oai_token-act-pair · gemini-2.0-flash

    phrases or tokens related to concluding thoughts or final answers in mathematical or logical reasoning problems.

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @john
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    fnlp/Llama-Scope-R1-Distill/400M-Slimpajama-400M-OpenR1-Math-220k/L11R
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Hzfinfdu/SlimPajama-3B and open-r1/OpenR1-Math-220k
    Features
    32,768
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    cerebras/SlimPajama-627B
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ican
    -0.06
    aliz
    -0.06
    èĭ
    -0.06
    опол
    -0.06
    æ±½
    -0.06
    incare
    -0.06
    urdy
    -0.06
    usan
    -0.06
    .WriteByte
    -0.06
    ¹
    -0.06
    POSITIVE LOGITS
     End
    0.07
     end
    0.07
     END
    0.07
    .END
    0.06
    lint
    0.06
    cec
    0.06
    End
    0.06
    834
    0.06
     endif
    0.06
     ends
    0.06
    Activations Density 0.100%

    No Known Activations