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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 31239
    Prev
    Next
    INDEX
    Explanations

    references to legal statutes and sections

    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
    258
    +0.17
    0.9%
    340
    +0.13
    0.8%
    271
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    66
    +0.17
    0.02
    82
    +0.13
    0.02
    37
    +0.12
    0.01
    Negative Logits
     contrary
    -1.45
     contempor
    -1.43
    stairs
    -1.42
    standing
    -1.42
    ]>
    -1.40
     veget
    -1.39
    lining
    -1.38
     close
    -1.38
     anybody
    -1.37
    local
    -1.36
    POSITIVE LOGITS
    Ĩ
    3.04
    ®
    2.48
    º
    2.43
    ´
    2.33
    ĸ´
    2.24
    ¿½
    2.23
    ¿
    2.17
    Į
    2.16
    Ĭ
    2.07
    §
    2.06
    Activations Density 0.079%

    No Known Activations