© 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. 2-RES-SM
    11. 31972
    Prev
    Next
    INDEX
    Explanations

    elements related to music and performance in a variety of settings

    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/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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.16
    0.9%
    16
    +0.12
    0.7%
    319
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    443
    +0.16
    0.03
    462
    +0.12
    0.02
    319
    +0.12
    0.03
    Negative Logits
     DOI
    -1.71
     NHS
    -1.65
     Privacy
    -1.63
     vaccine
    -1.62
    amacare
    -1.60
     Migration
    -1.60
     Site
    -1.58
    ){#
    -1.57
     Planned
    -1.56
     doi
    -1.55
    POSITIVE LOGITS
     music
    2.77
     songs
    2.54
     singing
    2.53
     sung
    2.53
     musical
    2.42
     bass
    2.41
     guitar
    2.41
     musicians
    2.36
     musician
    2.34
    music
    2.33
    Activations Density 0.628%

    No Known Activations