© 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. 32167
    Prev
    Next
    INDEX
    Explanations

    strings that represent country names and their associated identifiers

    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
    +0.30
    1.8%
    395
    +0.16
    0.9%
    5
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    395
    +0.30
    0.17
    510
    +0.16
    0.17
    387
    +0.13
    0.12
    Negative Logits
    ĥ½
    -2.01
    ¤
    -1.76
    quito
    -1.68
    acer
    -1.64
    Ĺ
    -1.51
    §
    -1.48
    ī
    -1.47
    coli
    -1.45
    ®
    -1.45
    ¿½
    -1.44
    POSITIVE LOGITS
    production
    1.44
    tering
    1.42
     wherein
    1.34
     HERE
    1.34
     (`
    1.29
     Writers
    1.26
     which
    1.26
     here
    1.25
    ).[
    1.25
     occupied
    1.24
    Activations Density 5.178%

    No Known Activations