© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. MLP Post
    5. 2-MLP-SM
    6. 26943
    Prev
    Next
    INDEX
    Explanations

    instances of the word "attached."

    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
    307
    +0.14
    0.8%
    327
    +0.12
    0.6%
    499
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    295
    +0.14
    0.04
    375
    +0.12
    0.03
    198
    +0.11
    0.03
    Negative Logits
    abad
    -1.56
    )</
    -1.55
    noreply
    -1.51
    )?
    -1.48
     Modi
    -1.47
    ollen
    -1.44
     offence
    -1.40
    edly
    -1.40
    )>
    -1.39
     thereof
    -1.39
    POSITIVE LOGITS
    adors
    1.59
    Ń
    1.53
    tico
    1.42
     ourselves
    1.42
     unpublished
    1.41
    erton
    1.39
    ĩ
    1.37
    dens
    1.33
    zet
    1.32
     [^
    1.32
    Activations Density 0.490%

    No Known Activations