© 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. 26260
    Prev
    Next
    INDEX
    Explanations

    expressions and themes related to love and emotional connections

    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
    56
    +0.14
    0.8%
    38
    +0.13
    0.8%
    232
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    38
    +0.14
    0.06
    56
    +0.13
    -0.00
    394
    +0.12
    0.08
    Negative Logits
    ĻĤ
    -2.37
    »¿
    -2.12
    ¿
    -1.89
    ames
    -1.74
    aucoup
    -1.61
    ·¸
    -1.58
    »
    -1.57
    inees
    -1.55
     sarcoma
    -1.55
    ĵ
    -1.55
    POSITIVE LOGITS
    ful
    3.19
    fulness
    2.75
    fully
    2.74
    lessness
    2.25
    ableness
    2.24
    ously
    2.12
    ingly
    2.09
    lessly
    2.06
     dissip
    1.90
    less
    1.88
    Activations Density 1.096%

    No Known Activations