© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 32183
    Prev
    Next
    INDEX
    Explanations

    references to physical office spaces and related environments

    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
    451
    +0.15
    0.8%
    251
    +0.13
    0.7%
    156
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    251
    +0.15
    0.03
    451
    +0.13
    0.02
    145
    +0.13
    0.03
    Negative Logits
    ĸ
    -2.40
    Į
    -2.24
    Ĵ
    -2.19
    Ĥ
    -2.16
    ³
    -2.00
    ı
    -1.99
    ĥ½
    -1.96
    Ī
    -1.94
     breath
    -1.92
    ĨĴ
    -1.92
    POSITIVE LOGITS
    holder
    2.33
    holders
    2.31
    yard
    2.01
    piece
    1.96
     personnel
    1.90
    yards
    1.89
     manager
    1.85
    inar
    1.84
     assistant
    1.83
     area
    1.81
    Activations Density 0.210%

    No Known Activations