© 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. 27704
    Prev
    Next
    INDEX
    Explanations

    phrases indicating conditions that may require medical attention or caution regarding the use of 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
    478
    +0.12
    0.7%
    35
    +0.12
    0.7%
    503
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    209
    +0.12
    0.10
    494
    +0.12
    0.09
    511
    +0.12
    0.08
    Negative Logits
    xit
    -1.72
    inent
    -1.68
    went
    -1.53
    ives
    -1.48
    ounding
    -1.47
    entre
    -1.46
    stown
    -1.45
    named
    -1.38
    ocene
    -1.36
    ãģ¦
    -1.36
    POSITIVE LOGITS
    naments
    2.56
    leans
    2.02
    chid
    2.00
     else
    1.91
     otherwise
    1.91
    IENT
    1.85
    deal
    1.83
     equivalently
    1.80
    ifice
    1.77
    bital
    1.69
    Activations Density 0.292%

    No Known Activations