© 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. 2-RES-SM
    11. 26608
    Prev
    Next
    INDEX
    Explanations

    references to programming constructs and operations in Swift

    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
    360
    +0.14
    0.8%
    65
    +0.14
    0.8%
    79
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    79
    +0.14
    0.04
    157
    +0.14
    0.05
    122
    +0.13
    0.03
    Negative Logits
     purposes
    -1.59
    illion
    -1.52
     pleading
    -1.51
    arro
    -1.48
    à±ģ
    -1.47
    oustic
    -1.42
    à°¿
    -1.40
    =[
    -1.35
    blic
    -1.34
    ending
    -1.34
    POSITIVE LOGITS
    °
    2.20
    ĺ
    1.89
     slack
    1.83
    ½
    1.69
    Īĺ
    1.68
    ¢
    1.66
     File
    1.60
    ©
    1.58
     mehr
    1.53
    ¾
    1.53
    Activations Density 0.505%

    No Known Activations