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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 28551
    Prev
    Next
    INDEX
    Explanations

    references to the concept of "doubling" or "double."

    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
    307
    +0.13
    0.7%
    196
    +0.12
    0.7%
    420
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    196
    +0.13
    0.02
    32
    +0.12
    0.02
    275
    +0.12
    0.01
    Negative Logits
    INE
    -1.64
    ĵ
    -1.61
    ĸ
    -1.61
    NOT
    -1.57
    using
    -1.55
    frastructure
    -1.54
    rapeut
    -1.48
    IRE
    -1.48
    astro
    -1.47
    unnumbered
    -1.47
    POSITIVE LOGITS
    heit
    2.07
    uple
    1.94
     sized
    1.89
    horn
    1.86
     entend
    1.86
     digits
    1.72
     dice
    1.56
    bow
    1.56
     jeopardy
    1.55
    ts
    1.48
    Activations Density 0.093%

    No Known Activations