© 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. 1-RES-SM
    11. 24099
    Prev
    Next
    INDEX
    Explanations

    references to flexibility in various contexts

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    +0.32
    1.8%
    376
    +0.15
    0.9%
    274
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    231
    +0.32
    0.02
    448
    +0.15
    0.02
    339
    +0.13
    0.02
    Negative Logits
    anan
    -1.49
     thrilling
    -1.49
     vain
    -1.49
    ylum
    -1.47
    inine
    -1.43
     shout
    -1.41
     cruel
    -1.41
     fright
    -1.39
     reckless
    -1.39
     falsely
    -1.39
    POSITIVE LOGITS
    £
    3.27
    ģ
    3.21
    ĻĤ
    3.12
    ĸ
    3.08
    ļ
    3.06
    ¢
    3.03
    Ļ
    3.01
    ł
    3.00
    ŀ
    2.99
    ķ
    2.93
    Activations Density 0.081%

    No Known Activations