© 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. 11409
    Prev
    Next
    INDEX
    Explanations

    possessive pronouns and indications of ownership

    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
    156
    +0.18
    1.0%
    478
    +0.14
    0.8%
    379
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    379
    +0.18
    0.08
    50
    +0.14
    0.04
    376
    +0.12
    0.01
    Negative Logits
    ioned
    -1.58
    âĦ¢
    -1.58
    II
    -1.54
    uously
    -1.54
    ously
    -1.53
    III
    -1.49
    ingu
    -1.47
    ificates
    -1.42
    FC
    -1.37
    shall
    -1.35
    POSITIVE LOGITS
    Ļª
    2.29
    ľĵ
    2.18
    »
    2.17
    ĻĤ
    2.12
    ±
    2.09
    ĭ
    2.08
    ĨĴ
    2.07
    ¹
    2.04
    Īĺ
    1.99
    ī
    1.96
    Activations Density 0.941%

    No Known Activations