© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 12232
    Prev
    Next
    INDEX
    Explanations

    interrogative sentences that pose questions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    302
    +0.14
    0.8%
    419
    +0.13
    0.7%
    138
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    302
    +0.14
    0.06
    16
    +0.13
    0.05
    390
    +0.13
    0.03
    Negative Logits
    Ļª
    -2.40
    ģ
    -2.32
    ļ
    -2.10
    Ĵ
    -2.02
    ĸ
    -1.98
    ¼
    -1.90
    ı
    -1.89
    ¨
    -1.88
    ŀ
    -1.87
    į
    -1.86
    POSITIVE LOGITS
     questions
    2.37
    questions
    2.21
     answered
    2.19
     aloud
    2.05
    answered
    2.02
    Answer
    1.90
    answer
    1.89
     Question
    1.89
     asking
    1.88
     answering
    1.80
    Activations Density 1.111%

    No Known Activations