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. Residuals
    5. 2-RES-SM
    6. 28861
    Prev
    Next
    INDEX
    Explanations

    market-related terms and concepts, particularly in the context of television and sports broadcasting

    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
    140
    +0.16
    0.9%
    44
    +0.15
    0.8%
    92
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    140
    +0.16
    0.11
    109
    +0.15
    0.09
    44
    +0.13
    0.09
    Negative Logits
    ""
    -1.45
    erto
    -1.44
    criptor
    -1.40
    definition
    -1.38
    ection
    -1.38
    ){
    -1.37
    ound
    -1.37
    "));
    -1.35
    ffic
    -1.34
    unnumbered
    -1.33
    POSITIVE LOGITS
    Īĺ
    2.31
    ģ
    2.15
    º
    2.15
    ®
    2.14
    µ
    2.08
    ĥ
    2.07
    ¿
    2.02
    ¹
    2.00
    £
    1.91
    ¾
    1.85
    Activations Density 1.259%

    No Known Activations