© 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. 14108
    Prev
    Next
    INDEX
    Explanations

    formatted data and structure code elements

    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
    230
    +0.13
    0.7%
    453
    +0.12
    0.7%
    136
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    453
    +0.13
    0.12
    85
    +0.12
    0.06
    167
    +0.11
    0.07
    Negative Logits
     called
    -1.41
     pyl
    -1.35
     <-
    -1.35
    >`
    -1.33
     simplest
    -1.33
     fry
    -1.33
     bott
    -1.31
    opically
    -1.29
    ?’
    -1.28
    amethasone
    -1.28
    POSITIVE LOGITS
    Ļª
    3.31
    ¼
    3.18
    ģ
    2.93
    Ľ
    2.93
    ½
    2.92
    ±
    2.92
    ĺ
    2.90
    Ĵ
    2.84
    ľĵ
    2.83
    ŀ
    2.81
    Activations Density 3.821%

    No Known Activations