© 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. 14068
    Prev
    Next
    INDEX
    Explanations

    references to age-related issues and decline

    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
    478
    +0.16
    0.9%
    136
    +0.13
    0.7%
    407
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    457
    +0.16
    0.10
    404
    +0.13
    0.10
    407
    +0.13
    0.10
    Negative Logits
    (@
    -1.65
    wide
    -1.46
     respects
    -1.44
    berg
    -1.43
     orders
    -1.40
    avier
    -1.38
    [:
    -1.36
    ring
    -1.35
    greg
    -1.33
    bach
    -1.32
    POSITIVE LOGITS
    fered
    1.68
    ¦
    1.60
     everyday
    1.56
    ľĵ
    1.52
    ĥ½
    1.49
    yland
    1.48
           ↵
    1.48
    ↵
    1.48
    ↵
    1.48
    <|outofrange|>
    1.48
    Activations Density 1.023%

    No Known Activations