© 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. 23845
    Prev
    Next
    INDEX
    Explanations

    mentions of biological activity or processes

    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
    156
    +0.22
    1.3%
    41
    +0.11
    0.7%
    376
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    230
    +0.22
    0.02
    41
    +0.11
    0.01
    135
    +0.10
    0.01
    Negative Logits
    deen
    -1.65
     romantic
    -1.56
     famous
    -1.55
     acquaintance
    -1.54
     MERCHANTABILITY
    -1.52
     say
    -1.50
    ±
    -1.49
    llll
    -1.49
     Acad
    -1.48
     numerous
    -1.47
    POSITIVE LOGITS
    ically
    1.94
    ivity
    1.93
    ively
    1.82
    etically
    1.76
    ives
    1.73
    ually
    1.72
    meter
    1.72
     accrued
    1.68
    ivities
    1.67
    clock
    1.61
    Activations Density 0.044%

    No Known Activations