© 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. 4-RES-SM
    11. 11401
    Prev
    Next
    INDEX
    Explanations

    mentions of "lung" or related terms in the context of health and medical conditions

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    275
    +0.16
    0.9%
    195
    +0.13
    0.7%
    189
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    275
    +0.16
    0.01
    195
    +0.13
    0.01
    164
    +0.12
    0.01
    Negative Logits
     things
    -1.54
    naments
    -1.53
    rist
    -1.52
    usement
    -1.50
    quot
    -1.47
     pace
    -1.43
    ="@+
    -1.42
     endless
    -1.41
     rapport
    -1.41
    antly
    -1.39
    POSITIVE LOGITS
     lobes
    2.21
     adenocarcinoma
    2.15
     parench
    2.09
     cancer
    2.09
    es
    2.06
     cancers
    2.05
     nodules
    1.97
     carcinoma
    1.91
     lobe
    1.87
     transplantation
    1.83
    Activations Density 0.029%

    No Known Activations