© 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. 2-RES-SM
    11. 32338
    Prev
    Next
    INDEX
    Explanations

    references to low-cost options or affordability

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    85
    +0.14
    0.8%
    369
    +0.14
    0.8%
    462
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    85
    +0.14
    0.02
    227
    +0.14
    0.03
    451
    +0.11
    0.02
    Negative Logits
    axel
    -1.73
    usiness
    -1.49
    endment
    -1.49
    unlock
    -1.48
    avage
    -1.40
     redes
    -1.38
    agons
    -1.36
    NOW
    -1.35
    UTERS
    -1.34
    ICO
    -1.34
    POSITIVE LOGITS
    ermost
    2.64
     (\<
    2.27
    ercase
    2.20
     enough
    2.09
    liness
    2.01
    -$
    1.96
     (>
    1.85
     degree
    1.82
    rance
    1.82
     (\>
    1.81
    Activations Density 0.091%

    No Known Activations