© 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. 1-RES-SM
    11. 25874
    Prev
    Next
    INDEX
    Explanations

    words related to the presence of hazardous substances

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    87
    +0.11
    0.6%
    51
    +0.11
    0.6%
    385
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.11
    0.03
    271
    +0.11
    0.00
    279
    +0.11
    0.02
    Negative Logits
    atur
    -1.68
     Anyway
    -1.48
     choice
    -1.47
     mammary
    -1.43
    ieurs
    -1.42
     bother
    -1.39
    looking
    -1.38
    )$)
    -1.33
     stomach
    -1.33
    pping
    -1.33
    POSITIVE LOGITS
    İ
    2.61
    Ĵ
    2.59
    ®
    2.38
    ¸
    2.32
    ļ
    2.27
    ĩ
    2.26
    ±
    2.24
    Ħ
    2.22
    °
    2.17
    ī
    2.11
    Activations Density 0.100%

    No Known Activations