Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. MLP Post
    5. 2-MLP-SM
    6. 26152
    Prev
    Next
    INDEX
    Explanations

    adjectives describing size, quality, and complexity in a text

    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
    15
    +0.14
    0.7%
    502
    +0.11
    0.6%
    423
    +0.10
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    225
    +0.14
    0.07
    15
    +0.11
    0.07
    502
    +0.10
    0.07
    Negative Logits
    ¨
    -2.16
    ª
    -2.00
    ij
    -1.93
    Ł
    -1.91
    ¥
    -1.82
    ses
    -1.79
    IGHT
    -1.75
    ¤
    -1.72
    µ
    -1.72
    ¸
    -1.71
    POSITIVE LOGITS
    yard
    1.68
    ferroni
    1.60
    phab
    1.56
    vier
    1.54
    golang
    1.49
    bird
    1.45
    fox
    1.45
     Page
    1.45
    appa
    1.44
    acia
    1.44
    Activations Density 0.696%

    No Known Activations