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

    Neuronpedia

    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. 10984
    Prev
    Next
    INDEX
    Explanations

    terms related to metabolic processes and their related factors

    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
    +0.20
    1.1%
    186
    +0.11
    0.6%
    29
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    304
    +0.20
    0.01
    29
    +0.11
    0.00
    207
    +0.11
    0.01
    Negative Logits
    owing
    -1.98
    aucoup
    -1.83
     unsuccessful
    -1.80
     else
    -1.79
    ying
    -1.72
     careful
    -1.70
     sincere
    -1.66
    yo
    -1.66
     frivolous
    -1.60
     hesitate
    -1.55
    POSITIVE LOGITS
    £
    3.37
    ¢
    3.17
    ĺ
    2.92
    ¤
    2.92
    ł
    2.90
    ·
    2.87
    Ĥ
    2.87
    ©
    2.79
    Ħ
    2.78
    ¨
    2.65
    Activations Density 0.029%

    No Known Activations