© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 11614
    Prev
    Next
    INDEX
    Explanations

    references to individuals, particularly focusing on occurrences of the word "person."

    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
    365
    +0.17
    0.9%
    456
    +0.11
    0.6%
    407
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    365
    +0.17
    0.03
    407
    +0.11
    0.03
    485
    +0.10
    0.02
    Negative Logits
    Ĭ
    -2.64
    ĵ
    -2.61
    ·¸
    -2.53
    ·
    -2.43
    ı
    -2.41
    -2.40
    ↵↵
    -2.40
    -2.40
    -2.40
    ↵
    -2.40
    POSITIVE LOGITS
    nel
    2.41
     who
    1.88
    ila
    1.86
    uscript
    1.81
    nal
    1.75
    iscus
    1.73
     owns
    1.70
    ager
    1.65
    acles
    1.65
    arman
    1.65
    Activations Density 0.185%

    No Known Activations