© 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. 15069
    Prev
    Next
    INDEX
    Explanations

    conjunctions and transitions in arguments

    oai_token-act-pair · gpt-4o-miniTriggered 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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    21
    +0.33
    2.0%
    316
    +0.13
    0.8%
    103
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    316
    +0.33
    0.08
    15
    +0.13
    0.10
    458
    +0.13
    0.04
    Negative Logits
    <|outofrange|>
    -10.32
    ↵
    -10.32
    ↵
    -10.32
    -10.32
    č↵
    -10.32
    ↵↵
    -10.32
    -10.32
    -10.32
         ↵
    -10.32
    ↵↵
    -10.32
    POSITIVE LOGITS
    ,
    1.79
    ;
    1.59
    ,)
    1.57
    mente
    1.51
     impractical
    1.46
     contrary
    1.45
     dil
    1.44
    >;
    1.43
    .—
    1.42
    *,*
    1.41
    Activations Density 0.985%

    No Known Activations