© 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. 2-RES-SM
    11. 11058
    Prev
    Next
    INDEX
    Explanations

    references to medical treatments and their effects

    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
    186
    +0.17
    0.9%
    369
    +0.12
    0.7%
    320
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    391
    +0.17
    0.09
    345
    +0.12
    0.08
    399
    +0.10
    0.06
    Negative Logits
    plemented
    -1.41
     audience
    -1.38
    located
    -1.35
    (@"
    -1.31
    gester
    -1.31
    ++)
    -1.29
    ++){
    -1.25
     Boolean
    -1.24
    respectively
    -1.24
     processor
    -1.23
    POSITIVE LOGITS
    Ĥ¬
    2.05
    »
    1.84
    ©
    1.79
    ·¸
    1.77
    Ł
    1.74
    ĩ
    1.72
    §
    1.71
    į
    1.65
    Ĥ
    1.64
    ħ
    1.62
    Activations Density 0.968%

    No Known Activations