© 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. 29933
    Prev
    Next
    INDEX
    Explanations

    references to specific identification numbers or IDs

    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
    350
    +0.14
    0.8%
    470
    +0.13
    0.8%
    326
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    470
    +0.14
    0.04
    350
    +0.13
    0.04
    95
    +0.13
    0.04
    Negative Logits
    áĢº
    -1.66
    IGHT
    -1.65
    rapeutics
    -1.56
    bbe
    -1.52
     full
    -1.49
     mercy
    -1.48
    ycin
    -1.46
     anything
    -1.44
    selves
    -1.43
     goodness
    -1.43
    POSITIVE LOGITS
    aho
    1.95
    ios
    1.81
    ="@+
    1.74
    irectional
    1.70
    estic
    1.70
    iot
    1.67
    ée
    1.66
     poses
    1.64
    Åij
    1.60
    yll
    1.59
    Activations Density 0.183%

    No Known Activations