© 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. 4-RES-SM
    11. 15123
    Prev
    Next
    INDEX
    Explanations

    phrases indicating responsibility and accountability

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    21
    +0.24
    1.4%
    279
    +0.13
    0.8%
    78
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    254
    +0.24
    0.02
    149
    +0.13
    0.08
    194
    +0.12
    0.05
    Negative Logits
    ↵↵
    -3.99
    č↵č↵
    -3.99
    -3.99
    -3.99
    č↵
    -3.99
     ↵
    -3.99
    -3.99
    ↵
    -3.99
    č↵č↵
    -3.99
    ↵
    -3.99
    POSITIVE LOGITS
     tricks
    1.88
     heels
    1.78
     throttle
    1.75
     punches
    1.60
     mercy
    1.58
     upside
    1.58
     estoppel
    1.57
     bait
    1.55
     damned
    1.54
     knots
    1.52
    Activations Density 1.828%

    No Known Activations