© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 26441
    Prev
    Next
    INDEX
    Explanations

    parenthetical expressions or interruptions in 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
    387
    +0.14
    0.8%
    352
    +0.12
    0.7%
    337
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    352
    +0.14
    0.04
    334
    +0.12
    0.06
    284
    +0.12
    0.04
    Negative Logits
    helm
    -1.56
    ishers
    -1.55
    irms
    -1.45
    ____________
    -1.44
    ake
    -1.43
    enter
    -1.43
    second
    -1.42
    rating
    -1.42
    cker
    -1.41
    gel
    -1.41
    POSITIVE LOGITS
     construction
    1.59
    )$-
    1.57
     unchanged
    1.51
    cology
    1.50
    ')\
    1.46
    ).\
    1.45
     translocation
    1.41
     procurement
    1.40
     synthesis
    1.38
    ology
    1.37
    Activations Density 0.304%

    No Known Activations