© 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. 4-RES-SM
    11. 8556
    Prev
    Next
    INDEX
    Explanations

    specific phrases or words related to community involvement or engagement

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Neuron Alignment
    Index
    Value
    % of L₁
    271
    +0.27
    1.5%
    64
    +0.19
    1.1%
    21
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    64
    +0.27
    0.04
    271
    +0.19
    0.01
    501
    +0.13
    0.01
    Negative Logits
    htra
    -1.74
    ional
    -1.65
    psin
    -1.59
    isation
    -1.54
    heed
    -1.53
    burgh
    -1.53
    pan
    -1.52
    olver
    -1.47
    icide
    -1.47
    ities
    -1.47
    POSITIVE LOGITS
    ı
    4.59
    Ĵ
    4.43
    Ļ
    4.32
    ³
    4.29
    ĨĴ
    4.28
    Ģ
    4.21
    ²
    4.21
    Ī
    4.17
    ¶
    4.16
    °
    4.15
    Activations Density 0.040%

    No Known Activations