© 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 EvalsExportsAPI 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. 11223
    Prev
    Next
    INDEX
    Explanations

    adverbs that describe manner or frequency

    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%
    397
    +0.15
    0.8%
    9
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    263
    +0.18
    0.09
    117
    +0.15
    0.08
    511
    +0.13
    0.08
    Negative Logits
    ·¸
    -6.53
    ĥ½
    -6.50
    č↵
    -6.37
    ↵
    -6.37
    -6.37
          ↵
    -6.37
    <|outofrange|>
    -6.37
              ↵
    -6.37
    <|outofrange|>
    -6.37
    ↵
    -6.37
    POSITIVE LOGITS
     thereafter
    2.08
     since
    2.07
     without
    2.04
     after
    1.99
     during
    1.99
     postoperatively
    1.95
     despite
    1.95
     throughout
    1.94
     though
    1.93
     but
    1.93
    Activations Density 1.001%

    No Known Activations