© 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. 1-RES-SM
    11. 23954
    Prev
    Next
    INDEX
    Explanations

    instances of the word "campaign" related to various forms of political or advocacy efforts

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.17
    1.0%
    295
    +0.14
    0.8%
    72
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    242
    +0.17
    0.03
    72
    +0.14
    0.03
    295
    +0.13
    0.02
    Negative Logits
    ĥ½
    -2.64
    ²
    -2.57
    Ĵ
    -2.29
    ª
    -2.24
    ¯
    -2.14
    Į
    -2.09
    ħ
    -2.08
    ķ
    -2.04
    º
    -2.03
    ĭ
    -2.03
    POSITIVE LOGITS
    ers
    2.00
    master
    1.99
    holder
    1.96
    raising
    1.82
    chair
    1.82
    inator
    1.74
    holders
    1.73
    ulses
    1.72
    eries
    1.70
    tees
    1.69
    Activations Density 0.160%

    No Known Activations