© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 14994
    Prev
    Next
    INDEX
    Explanations

    references to books and related terms

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    410
    +0.12
    0.7%
    263
    +0.12
    0.6%
    10
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    38
    +0.12
    0.03
    263
    +0.12
    0.04
    10
    +0.11
    0.03
    Negative Logits
    ATED
    -1.70
    egal
    -1.60
    arily
    -1.50
     cysts
    -1.45
     turf
    -1.44
    omonas
    -1.42
    ullary
    -1.42
    ial
    -1.39
    aeda
    -1.39
    ESULT
    -1.36
    POSITIVE LOGITS
    worm
    3.03
    marks
    3.02
    mark
    2.84
    shop
    2.73
    store
    2.71
    keeping
    2.55
    marked
    2.54
    ends
    2.49
    keeper
    2.45
    case
    2.40
    Activations Density 0.263%

    No Known Activations