© 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. 15085
    Prev
    Next
    INDEX
    Explanations

    phrases that express logical reasoning or qualifications

    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/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
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    156
    +0.18
    1.0%
    289
    +0.14
    0.8%
    83
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    289
    +0.18
    0.08
    83
    +0.14
    0.10
    458
    +0.13
    0.04
    Negative Logits
    ĻĤ
    -2.14
    Ħ
    -1.99
    dAtA
    -1.88
    »¿
    -1.87
    áĢIJ
    -1.77
    Į
    -1.77
    "}](#
    -1.72
    ģ
    -1.71
    Ĭ
    -1.71
    áĢº
    -1.71
    POSITIVE LOGITS
     worth
    2.40
     regarded
    2.22
     indicative
    2.18
     compatible
    2.15
     amenable
    2.10
     preferable
    2.09
     capable
    2.08
     susceptible
    2.04
     dependent
    2.02
     considered
    1.97
    Activations Density 0.749%

    No Known Activations