© 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. 14151
    Prev
    Next
    INDEX
    Explanations

    bacterial names and their associated classifications

    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
    111
    +0.13
    0.8%
    397
    +0.13
    0.7%
    497
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.13
    0.00
    352
    +0.13
    0.02
    399
    +0.13
    0.01
    Negative Logits
    arning
    -2.10
      ↵
    -1.72
    aptic
    -1.64
    eting
    -1.61
    JUST
    -1.60
    ographics
    -1.56
     '</
    -1.55
    boards
    -1.51
    vez
    -1.49
    ).$$
    -1.47
    POSITIVE LOGITS
     pylori
    2.02
     aeruginosa
    1.96
    obacterium
    1.92
     strains
    1.88
     itself
    1.81
     aureus
    1.78
     cerevisiae
    1.78
    ococcus
    1.76
     spp
    1.74
     pneumoniae
    1.70
    Activations Density 0.123%

    No Known Activations