© 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. 4-RES-SM
    11. 12278
    Prev
    Next
    INDEX
    Explanations

    legal terminology and references related to judicial opinions

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    78
    +0.15
    0.9%
    378
    +0.12
    0.7%
    216
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    216
    +0.15
    0.02
    367
    +0.12
    0.02
    256
    +0.11
    0.01
    Negative Logits
    npmjs
    -1.76
    ries
    -1.56
    liest
    -1.54
     behaviour
    -1.50
    afa
    -1.45
    rapy
    -1.44
    rapeut
    -1.42
    ander
    -1.41
    avi
    -1.40
    orest
    -1.39
    POSITIVE LOGITS
    ģ
    2.64
    ¸
    2.60
    «
    2.37
    ²
    2.18
    ®
    2.18
    ¶
    2.13
    ¥
    2.13
    Ľ
    2.12
    ª
    2.11
     AND
    2.07
    Activations Density 0.057%

    No Known Activations