© 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. 1-RES-SM
    11. 25983
    Prev
    Next
    INDEX
    Explanations

    punctuations or end-of-sentence markers

    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
    135
    +0.21
    1.2%
    292
    +0.14
    0.8%
    508
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    135
    +0.21
    0.07
    508
    +0.14
    0.06
    43
    +0.13
    0.06
    Negative Logits
     collapse
    -1.72
     deceased
    -1.51
     bite
    -1.46
    oting
    -1.44
     mistrial
    -1.42
     mutants
    -1.37
     memoir
    -1.34
    collapse
    -1.34
     close
    -1.33
     failure
    -1.32
    POSITIVE LOGITS
    oft
    1.68
    ingale
    1.62
    osph
    1.53
    ÅĦ
    1.52
    dala
    1.51
    chal
    1.50
    chus
    1.50
    uvre
    1.50
    stract
    1.45
    enix
    1.45
    Activations Density 1.248%

    No Known Activations