© 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. 27803
    Prev
    Next
    INDEX
    Explanations

    mentions of academic credentials and professional affiliations

    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
    340
    +0.11
    0.6%
    398
    +0.11
    0.6%
    385
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.11
    0.12
    156
    +0.11
    0.15
    204
    +0.10
    0.12
    Negative Logits
    "}](#
    -1.71
    ricane
    -1.66
    ricanes
    -1.62
    ients
    -1.58
    lectric
    -1.54
     shutdown
    -1.50
    otropic
    -1.49
    abouts
    -1.47
    rians
    -1.46
    cted
    -1.44
    POSITIVE LOGITS
     manuscript
    2.61
     reprinted
    2.55
     editions
    2.52
     manuscripts
    2.50
     edition
    2.35
     autobi
    2.28
     illustrations
    2.28
     bibli
    2.26
     literary
    2.24
     chapters
    2.15
    Activations Density 4.012%

    No Known Activations