© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 10028
    Prev
    Next
    INDEX
    Explanations

    JSON-like structured data attributes and their values

    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
    156
    +0.23
    1.3%
    17
    +0.14
    0.8%
    397
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    17
    +0.23
    0.08
    397
    +0.14
    0.07
    325
    +0.13
    0.10
    Negative Logits
    hores
    -1.73
    '):
    -1.63
    yrinth
    -1.48
    sto
    -1.47
     urge
    -1.44
    ’,
    -1.44
    ’.
    -1.41
    ULTS
    -1.41
    gap
    -1.39
    lemental
    -1.38
    POSITIVE LOGITS
    ¦
    3.66
    ĥ½
    3.63
      ↵
    3.51
    3.51
    č↵č↵
    3.51
    ↵
    3.51
    3.51
    3.51
    ↵
    3.51
    ↵
    3.51
    Activations Density 1.122%

    No Known Activations