© 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. 29720
    Prev
    Next
    INDEX
    Explanations

    second-person pronouns and references to the reader

    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
    369
    +0.17
    0.9%
    478
    +0.13
    0.7%
    159
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    363
    +0.17
    0.04
    159
    +0.13
    0.05
    260
    +0.12
    0.04
    Negative Logits
    Ļª
    -4.23
            ↵
    -3.70
    -3.70
    <|outofrange|>
    -3.70
    ↵
    -3.70
    ↵
    -3.70
    ↵↵
    -3.70
          ↵
    -3.70
    <|outofrange|>
    -3.70
    -3.70
    POSITIVE LOGITS
     guys
    1.92
     personally
    1.78
     pocket
    1.74
    'll
    1.71
    fts
    1.69
    're
    1.63
    pg
    1.60
    pherd
    1.51
     bet
    1.45
    wl
    1.45
    Activations Density 0.268%

    No Known Activations