© 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. MLP Post
    9. 2-MLP-SM
    11. 24910
    Prev
    Next
    INDEX
    Explanations

    references to air and its qualities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/2-mlp-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_mlp_out
    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
    156
    +0.25
    1.5%
    23
    +0.21
    1.2%
    139
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    139
    +0.25
    0.03
    23
    +0.21
    0.01
    68
    +0.14
    0.02
    Negative Logits
    Ł
    -3.22
    ¼
    -3.18
    º
    -3.04
    ↵↵
    -3.00
    ↵
    -3.00
      ↵
    -3.00
    -3.00
    -3.00
    ↵↵
    -3.00
    ↵
    -3.00
    POSITIVE LOGITS
     ')
    1.56
     ")
    1.54
    ching
    1.52
    **)
    1.42
     understand
    1.41
     behold
    1.40
     perturbed
    1.31
    **
    1.31
     **
    1.30
    compare
    1.30
    Activations Density 0.192%

    No Known Activations