© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 14414
    Prev
    Next
    INDEX
    Explanations

    dietary recommendations and nutritional information

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.17
    1.0%
    29
    +0.11
    0.6%
    328
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    328
    +0.17
    0.03
    56
    +0.11
    -0.00
    361
    +0.10
    0.02
    Negative Logits
     forth
    -2.17
    ķ
    -1.68
     assass
    -1.68
     tunneling
    -1.62
    ureus
    -1.60
    Ŀ
    -1.59
     cref
    -1.58
    ons
    -1.51
    hots
    -1.49
    µ
    -1.48
    POSITIVE LOGITS
    etic
    2.40
    etics
    1.81
    iversity
    1.69
     intake
    1.62
    calc
    1.61
     Nutrition
    1.60
    entary
    1.59
    iverse
    1.58
    vary
    1.55
    Food
    1.53
    Activations Density 0.196%

    No Known Activations