© 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 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. 9814
    Prev
    Next
    INDEX
    Explanations

    programming and technical terminology related to installation and system configuration

    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
    156
    +0.25
    1.4%
    369
    +0.17
    1.0%
    376
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    369
    +0.25
    0.05
    376
    +0.17
    0.02
    90
    +0.15
    0.07
    Negative Logits
    ypse
    -1.55
    ptions
    -1.46
    ipes
    -1.45
    stitutional
    -1.44
     worden
    -1.43
     gas
    -1.42
    elles
    -1.35
    аг
    -1.29
    afers
    -1.29
    iforn
    -1.28
    POSITIVE LOGITS
     Illustr
    1.60
     Size
    1.54
     Recovery
    1.52
     Build
    1.52
     Weekly
    1.51
    ought
    1.48
    uet
    1.45
     Week
    1.43
    lli
    1.42
    fileID
    1.42
    Activations Density 1.374%

    No Known Activations