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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 12058
    Prev
    Next
    INDEX
    Explanations

    phrases related to significant political events or legislative actions

    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
    186
    +0.13
    0.7%
    410
    +0.12
    0.7%
    270
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    418
    +0.13
    -0.05
    186
    +0.12
    0.06
    178
    +0.12
    0.10
    Negative Logits
    Ĵ
    -2.57
    Ł
    -2.42
    Ĭ
    -2.33
    ĥ½
    -2.20
    ¢
    -2.17
    Ĥ
    -2.13
    <|outofrange|>
    -2.06
    <|outofrange|>
    -2.06
    ↵                                             
    -2.06
                                           
    -2.06
    POSITIVE LOGITS
     Republicans
    3.27
     Republican
    3.13
     Obama
    3.12
     Democrats
    3.09
     Trump
    2.94
     congressional
    2.94
     Biden
    2.94
     GOP
    2.80
     bipartisan
    2.79
     Senate
    2.64
    Activations Density 5.215%

    No Known Activations