© 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. 4-RES-SM
    11. 11770
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate changes or outcomes related to health conditions or treatments

    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
    78
    +0.15
    0.8%
    474
    +0.11
    0.6%
    445
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    55
    +0.15
    0.08
    402
    +0.11
    0.08
    256
    +0.10
    0.07
    Negative Logits
    Ļª
    -5.20
    µ
    -4.81
    ľ
    -4.72
    ¨
    -4.71
    ↵
    -4.70
    -4.70
    ↵
    -4.70
              ↵
    -4.70
    -4.70
     ↵
    -4.70
    POSITIVE LOGITS
     been
    3.21
     arisen
    2.42
    been
    2.28
     become
    2.28
     emerged
    2.25
    hing
    2.21
     existed
    2.14
     implications
    2.12
     relevance
    2.07
     begun
    2.02
    Activations Density 0.229%

    No Known Activations