© 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. 1-RES-SM
    11. 23967
    Prev
    Next
    INDEX
    Explanations

    phrases indicating acknowledgment or validation

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    100
    +0.14
    0.8%
    156
    +0.13
    0.8%
    462
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    38
    +0.14
    0.01
    100
    +0.13
    0.02
    462
    +0.10
    0.01
    Negative Logits
    ĻĤ
    -2.82
    ½
    -2.55
    ¹
    -2.46
    ĥ½
    -2.46
    IJ
    -2.29
    ¿½
    -2.22
    ı
    -2.22
    Īĺ
    -2.21
    Ľ
    -2.12
    ¯
    -2.11
    POSITIVE LOGITS
     purposes
    1.80
     maker
    1.79
     recognition
    1.72
     instance
    1.69
     example
    1.66
     event
    1.65
     sequence
    1.64
     identification
    1.64
     card
    1.61
     detection
    1.59
    Activations Density 0.101%

    No Known Activations