© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 2-RES-SM
    11. 30103
    Prev
    Next
    INDEX
    Explanations

    references and notes in a document

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    223
    +0.13
    0.7%
    71
    +0.13
    0.7%
    329
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    329
    +0.13
    0.01
    223
    +0.13
    0.01
    131
    +0.12
    0.02
    Negative Logits
     majority
    -1.82
    Ł
    -1.73
    ľĵ
    -1.66
    ī
    -1.59
    ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ
    -1.49
     enough
    -1.48
    ¶
    -1.48
    ipes
    -1.47
    ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ
    -1.45
    ı
    -1.44
    POSITIVE LOGITS
    ↵↵ ↵
    1.74
    ↵↵
    1.66
    ↵ ↵
    1.61
    ↵ ↵
    1.55
     Between
    1.52
    aid
    1.45
     links
    1.41
    ↵ ↵ ↵ ↵
    1.37
     map
    1.35
     Were
    1.35
    Activations Density 0.054%

    No Known Activations