© 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. 10442
    Prev
    Next
    INDEX
    Explanations

    phrases that highlight performance metrics and comparisons

    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
    123
    +0.12
    0.7%
    196
    +0.11
    0.6%
    158
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    158
    +0.12
    0.03
    196
    +0.11
    0.02
    462
    +0.11
    0.02
    Negative Logits
    ¼
    -2.42
    º
    -2.00
    mina
    -1.95
    ´
    -1.91
    jours
    -1.90
    ¿½
    -1.87
    Ĥ
    -1.81
    ª
    -1.80
    ĻĤ
    -1.79
    »
    -1.77
    POSITIVE LOGITS
     efficiency
    1.69
     results
    1.65
     indicators
    1.62
     enhancing
    1.61
     measures
    1.57
     according
    1.56
    ont
    1.51
     efficiencies
    1.51
     deficits
    1.50
    etric
    1.49
    Activations Density 0.097%

    No Known Activations