© 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. 14156
    Prev
    Next
    INDEX
    Explanations

    references to specific zones, particularly in geographical or technical contexts

    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
    30
    +0.15
    0.9%
    156
    +0.13
    0.8%
    115
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    30
    +0.15
    0.02
    115
    +0.13
    0.02
    413
    +0.11
    0.01
    Negative Logits
     pardon
    -1.62
     promise
    -1.57
     speaking
    -1.55
     oath
    -1.48
     utter
    -1.47
    terday
    -1.45
    looking
    -1.41
    MENT
    -1.37
     jus
    -1.36
     awful
    -1.36
    POSITIVE LOGITS
    illas
    2.15
    uto
    2.03
    oslav
    1.97
    ash
    1.92
    ipline
    1.90
    ucht
    1.86
    icki
    1.85
    ika
    1.77
    vier
    1.76
    illa
    1.76
    Activations Density 0.246%

    No Known Activations