© 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. 15025
    Prev
    Next
    INDEX
    Explanations

    terms related to legal or regulatory contexts

    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
    25
    +0.21
    1.3%
    156
    +0.21
    1.3%
    466
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    25
    +0.21
    0.17
    466
    +0.21
    0.12
    462
    +0.15
    0.06
    Negative Logits
    idenote
    -1.67
     unpublished
    -1.32
    rich
    -1.29
    Ģ
    -1.25
     speculate
    -1.23
     fois
    -1.23
    ème
    -1.21
     decir
    -1.21
     though
    -1.20
    orie
    -1.20
    POSITIVE LOGITS
    ↵
    1.87
    Âł↵
    1.63
    ↵Č
    1.51
    aped
    1.35
    ↵
    1.32
    Chapter
    1.28
    ...](
    1.26
     themselves
    1.25
     wrought
    1.23
      ↵
    1.23
    Activations Density 3.939%

    No Known Activations