© 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. 11093
    Prev
    Next
    INDEX
    Explanations

    phrases and terminology related to organizational structure and identification in content

    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
    302
    +0.18
    1.0%
    240
    +0.14
    0.8%
    218
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    140
    +0.18
    0.07
    218
    +0.14
    0.09
    302
    +0.12
    0.07
    Negative Logits
    à¯
    -1.67
    àµ
    -1.65
    àµį
    -1.58
    ãģ¾ãģĽãĤĵ
    -1.57
    ãģ¾ãģĹãģŁ
    -1.56
    à°¿
    -1.55
    à¯ģ
    -1.53
    à±ģ
    -1.53
    today
    -1.51
    ãģĹãģ¾ãģĻ
    -1.50
    POSITIVE LOGITS
    ↵³³
    1.65
     ground
    1.59
     unpublished
    1.49
     complex
    1.46
     record
    1.38
     Nobel
    1.38
     adversary
    1.37
     states
    1.36
     affiliations
    1.35
     realm
    1.34
    Activations Density 3.453%

    No Known Activations