© 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 EvalsExportsinterp-engineNEWAPI 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. 24337
    Prev
    Next
    INDEX
    Explanations

    references to various health and safety programs or initiatives

    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
    156
    +0.20
    1.2%
    498
    +0.15
    0.9%
    356
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    498
    +0.20
    0.04
    356
    +0.15
    0.03
    141
    +0.13
    0.02
    Negative Logits
    ĥ½
    -4.52
    ¥
    -3.96
    ĨĴ
    -3.65
    Ĵ
    -3.59
    ¬
    -3.54
    Ļª
    -3.48
    ¶
    -3.41
    ↵↵
    -3.38
    <|outofrange|>
    -3.38
    ↵
    -3.38
    POSITIVE LOGITS
    matic
    2.66
    emaker
    1.90
    matically
    1.79
    ic
    1.68
     aimed
    1.67
     focused
    1.63
    go
    1.63
    wide
    1.59
    irm
    1.58
    uri
    1.58
    Activations Density 0.134%

    No Known Activations