© 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. MLP Post
    9. 2-MLP-SM
    11. 22332
    Prev
    Next
    INDEX
    Explanations

    characteristics or qualities related to softness

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    23
    +0.26
    1.5%
    156
    +0.16
    0.9%
    193
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    23
    +0.26
    0.02
    511
    +0.16
    0.01
    374
    +0.12
    0.01
    Negative Logits
    chle
    -1.52
    notation
    -1.49
     forma
    -1.46
     figur
    -1.45
    culus
    -1.45
     modo
    -1.44
    posium
    -1.44
    icile
    -1.43
     Contest
    -1.43
    ctions
    -1.42
    POSITIVE LOGITS
     thick
    1.68
    iculously
    1.65
     \|
    1.62
     dense
    1.59
    ĻĤ
    1.56
    »¿
    1.52
     overwhelming
    1.51
     open
    1.49
     clear
    1.48
    ±
    1.47
    Activations Density 0.363%

    No Known Activations