© 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. Residuals
    9. 2-RES-SM
    11. 29939
    Prev
    Next
    INDEX
    Explanations

    keywords related to news articles and media organizations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-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_resid_post
    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
    302
    +0.14
    0.8%
    248
    +0.12
    0.7%
    57
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    64
    +0.14
    0.10
    204
    +0.12
    0.12
    418
    +0.11
    -0.05
    Negative Logits
    InstanceState
    -1.69
    ""
    -1.65
     stones
    -1.63
     whilst
    -1.60
     individually
    -1.57
     {}
    -1.53
     """
    -1.52
    ço
    -1.47
     otherwise
    -1.46
     however
    -1.45
    POSITIVE LOGITS
     newsletter
    2.14
     Tribune
    1.95
     comments
    1.91
     reporters
    1.90
     editorial
    1.82
     reporter
    1.71
     reprinted
    1.69
     headline
    1.69
     rebut
    1.68
     ESPN
    1.66
    Activations Density 4.373%

    No Known Activations