© 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 EvalsExportsinterp-engineNEWAPI 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. 24169
    Prev
    Next
    INDEX
    Explanations

    patterns involving numerical values and technical components

    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
    208
    +0.11
    0.6%
    411
    +0.10
    0.6%
    181
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    336
    +0.11
    0.04
    291
    +0.10
    0.03
    492
    +0.10
    0.03
    Negative Logits
    ipl
    -1.69
    panic
    -1.65
     *_
    -1.59
    *^
    -1.57
    ^\*^
    -1.56
    ^(
    -1.54
    ĻĤ
    -1.50
     \$
    -1.48
    (_
    -1.47
    §
    -1.43
    POSITIVE LOGITS
     ensued
    1.50
     finance
    1.41
     Aug
    1.34
    taire
    1.32
    rendum
    1.30
    lem
    1.25
     Arabia
    1.24
    ikel
    1.22
     councils
    1.22
     unanim
    1.22
    Activations Density 0.708%

    No Known Activations