© 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. 15283
    Prev
    Next
    INDEX
    Explanations

    language associated with safety concerns and its implications

    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
    +0.28
    1.7%
    156
    +0.24
    1.4%
    269
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    269
    +0.28
    0.13
    401
    +0.24
    0.04
    186
    +0.13
    0.09
    Negative Logits
    <|padding|>
    -5.25
    ↵
    -5.25
    <|outofrange|>
    -5.25
    ↵
    -5.25
    ↵
    -5.25
    ↵
    -5.25
    ↵
    -5.25
    -5.25
    <|outofrange|>
    -5.25
    ↵        ↵
    -5.25
    POSITIVE LOGITS
     nowadays
    1.84
     sometimes
    1.84
     usually
    1.83
     {#
    1.76
     typically
    1.75
    mith
    1.70
     often
    1.63
     everywhere
    1.62
     generally
    1.59
    ystems
    1.58
    Activations Density 2.222%

    No Known Activations