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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 1-RES-SM
    11. 25149
    Prev
    Next
    INDEX
    Explanations

    instances of increment operations in programming code

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    488
    +0.13
    0.7%
    156
    +0.11
    0.6%
    471
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.13
    0.03
    298
    +0.11
    0.02
    156
    +0.11
    0.02
    Negative Logits
    ·
    -4.00
    ľ
    -3.94
    ¹
    -3.92
    ³
    -3.85
    ¯
    -3.81
    ¦
    -3.77
    ´
    -3.76
    ¨
    -3.75
    ķ
    -3.75
    ł
    -3.71
    POSITIVE LOGITS
    etica
    1.66
     introdu
    1.63
    ű
    1.57
    cores
    1.55
    abad
    1.54
     expansions
    1.51
    ronic
    1.50
     conventions
    1.50
     dialect
    1.50
     polÃŃ
    1.45
    Activations Density 0.165%

    No Known Activations