© 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. 14302
    Prev
    Next
    INDEX
    Explanations

    terms related to ideologies or movements, particularly those involving specific political or social affiliations

    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/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
    Neuron Alignment
    Index
    Value
    % of L₁
    334
    +0.13
    0.7%
    22
    +0.11
    0.6%
    434
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    334
    +0.13
    0.03
    474
    +0.11
    0.02
    325
    +0.10
    0.02
    Negative Logits
    ´
    -1.95
    ¶
    -1.70
    )."
    -1.63
    ]."
    -1.57
    ·¸
    -1.56
    ®
    -1.56
    µ
    -1.51
    ĥ½
    -1.49
     Touch
    -1.47
     )"
    -1.46
    POSITIVE LOGITS
    ribution
    2.05
    rict
    1.88
    ributed
    1.86
    enum
    1.84
    ributions
    1.75
    ulus
    1.74
    unnel
    1.72
    emas
    1.70
    inct
    1.69
    ribute
    1.67
    Activations Density 0.134%

    No Known Activations