© 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. 2-RES-SM
    11. 30854
    Prev
    Next
    INDEX
    Explanations

    references to different iterations or variations of products, processes, or concepts

    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
    156
    +0.21
    1.2%
    33
    +0.14
    0.8%
    15
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    15
    +0.21
    0.07
    306
    +0.14
    0.05
    32
    +0.11
    0.03
    Negative Logits
    ¼
    -3.42
    §
    -3.41
    ¥
    -3.40
    Ĵ
    -3.20
    ı
    -3.13
    ¾
    -3.02
    ½
    -2.99
    ±
    -2.96
    ĥ½
    -2.95
    Ľ
    -2.95
    POSITIVE LOGITS
    hip
    2.32
    hell
    2.29
    cript
    2.29
    ional
    2.28
    ystems
    2.25
    mith
    2.22
    heet
    2.21
    icle
    2.15
    hips
    2.10
    ymbol
    2.04
    Activations Density 0.637%

    No Known Activations