© 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. 2-RES-SM
    11. 32330
    Prev
    Next
    INDEX
    Explanations

    programming and technical terms related to data management and analysis

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    271
    +0.19
    1.1%
    507
    +0.18
    1.1%
    293
    +0.16
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.19
    0.20
    156
    +0.18
    0.22
    71
    +0.16
    0.22
    Negative Logits
    :`
    -1.81
    itarian
    -1.48
     standpoint
    -1.48
     influence
    -1.47
     competition
    -1.42
     issue
    -1.39
    oretic
    -1.38
     niche
    -1.37
     perspective
    -1.36
     basis
    -1.34
    POSITIVE LOGITS
    ĻĤ
    2.22
    substack
    2.00
    mediately
    1.83
    ¦
    1.80
    ¤
    1.79
    Ĥ
    1.61
    ând
    1.60
    ickets
    1.58
    intent
    1.57
    Ŀ
    1.57
    Activations Density 4.717%

    No Known Activations