© 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. 32017
    Prev
    Next
    INDEX
    Explanations

    references to technical specifications and advancements in technology

    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
    44
    +0.13
    0.7%
    231
    +0.12
    0.7%
    266
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    266
    +0.13
    0.14
    44
    +0.12
    0.15
    176
    +0.11
    0.18
    Negative Logits
    Į
    -3.17
    ļ
    -3.01
    ¯
    -2.95
    Ĥ
    -2.92
    ¢
    -2.88
    ħ
    -2.87
    Ł
    -2.83
    ij
    -2.81
    ¾
    -2.78
    ¿½
    -2.76
    POSITIVE LOGITS
     stakes
    1.77
    rive
    1.60
     worthwhile
    1.48
    ()))
    1.38
    .<
    1.38
     whatsoever
    1.37
    starter
    1.35
    enic
    1.34
    .[]{
    1.34
     niche
    1.32
    Activations Density 8.555%

    No Known Activations