© 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. 1-RES-SM
    11. 24013
    Prev
    Next
    INDEX
    Explanations

    adverbs related to efficiency and effectiveness

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    23
    +0.13
    0.8%
    86
    +0.13
    0.7%
    156
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    86
    +0.13
    0.03
    377
    +0.13
    0.03
    257
    +0.12
    0.02
    Negative Logits
    aurus
    -1.97
    ones
    -1.85
    acious
    -1.73
    mania
    -1.68
    icans
    -1.67
    iere
    -1.64
     editions
    -1.63
     edition
    -1.59
    blic
    -1.59
    ħ
    -1.58
    POSITIVE LOGITS
     discharged
    1.64
     shutting
    1.43
     transferring
    1.42
     treating
    1.42
     Thomson
    1.41
    come
    1.35
     enough
    1.34
     metabol
    1.34
     shutdown
    1.34
     neutralizing
    1.34
    Activations Density 0.082%

    No Known Activations