© 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 EvalsExportsAPI 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. 10647
    Prev
    Next
    INDEX
    Explanations

    terms related to molecules and their interactions

    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
    144
    +0.15
    0.8%
    156
    +0.12
    0.7%
    467
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    144
    +0.15
    0.01
    8
    +0.12
    0.01
    453
    +0.11
    0.01
    Negative Logits
    ĥ½
    -1.75
     sale
    -1.66
    ril
    -1.63
     morning
    -1.60
    woke
    -1.53
    ŀ
    -1.52
     Marriage
    -1.51
     morn
    -1.51
    ahoma
    -1.50
    blogger
    -1.48
    POSITIVE LOGITS
    etically
    1.84
    iverse
    1.84
    olecule
    1.67
    emat
    1.63
    orphic
    1.61
    cules
    1.61
    orphisms
    1.54
    ault
    1.51
    wright
    1.51
    chie
    1.50
    Activations Density 0.081%

    No Known Activations