© 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. 0-MLP-SM
    11. 27672
    Prev
    Next
    INDEX
    Explanations

    instances of the word "comprises."

    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/0-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.0.hook_mlp_out
    Hook Layer
    0
    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
    376
    +0.21
    1.2%
    23
    +0.18
    1.0%
    115
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    37
    +0.21
    0.01
    316
    +0.18
    0.01
    266
    +0.13
    0.01
    Negative Logits
    ÅĽÄĩ
    -1.58
    ÃŃst
    -1.53
    istics
    -1.48
     served
    -1.46
    amento
    -1.42
    edo
    -1.41
    ÅĽci
    -1.41
    's
    -1.41
    books
    -1.40
     cited
    -1.39
    POSITIVE LOGITS
    imal
    1.63
     veins
    1.53
    inib
    1.50
    >=
    1.41
     pity
    1.41
    imet
    1.39
     inability
    1.39
    kill
    1.38
    omit
    1.38
    dimethyl
    1.37
    Activations Density 0.030%

    No Known Activations