© 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. 8787
    Prev
    Next
    INDEX
    Explanations

    various forms of the word 'Azure', often in the context of computing or programming

    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
    248
    +0.13
    0.7%
    245
    +0.11
    0.6%
    67
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.13
    0.03
    397
    +0.11
    0.03
    203
    +0.11
    0.02
    Negative Logits
     Robin
    -1.57
     Shaw
    -1.47
    inder
    -1.45
    arie
    -1.44
    ivity
    -1.40
     tall
    -1.38
    pector
    -1.36
     mercy
    -1.35
    Willi
    -1.33
    erals
    -1.32
    POSITIVE LOGITS
     Ac
    3.54
    ac
    3.39
    Ac
    3.07
    AC
    2.94
     ac
    2.86
     AC
    2.84
    acs
    2.70
     ACP
    2.61
    aco
    2.38
    aci
    2.36
    Activations Density 0.493%

    No Known Activations