© 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. 32510
    Prev
    Next
    INDEX
    Explanations

    weather-related data and measurements

    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
    326
    +0.17
    1.0%
    263
    +0.13
    0.8%
    220
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    220
    +0.17
    0.05
    326
    +0.13
    0.04
    294
    +0.13
    0.05
    Negative Logits
     continues
    -1.65
     provides
    -1.52
     industries
    -1.45
     together
    -1.42
     *)
    -1.42
     deserve
    -1.42
     entit
    -1.41
     represents
    -1.40
     profession
    -1.40
     themselves
    -1.39
    POSITIVE LOGITS
     someday
    1.57
     Airport
    1.47
     sight
    1.38
    sson
    1.36
     premature
    1.32
     gallon
    1.31
     curled
    1.30
    ipeg
    1.30
    abouts
    1.27
    schem
    1.25
    Activations Density 3.744%

    No Known Activations