© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 12367
    Prev
    Next
    INDEX
    Explanations

    legal terminology and references in various contexts

    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
    410
    +0.15
    0.8%
    492
    +0.14
    0.8%
    412
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.15
    0.08
    66
    +0.14
    0.06
    338
    +0.12
    0.03
    Negative Logits
    fections
    -2.21
    minded
    -1.85
    fected
    -1.62
    fection
    -1.60
    nel
    -1.48
     victims
    -1.46
     nervous
    -1.41
    clamation
    -1.41
     killers
    -1.41
     owners
    -1.40
    POSITIVE LOGITS
    ĻĤ
    2.54
    º
    2.13
    estock
    2.10
    £
    2.09
    Ļª
    2.09
    ¥
    2.05
    orentz
    2.00
    ª
    1.90
    horn
    1.87
    ³
    1.82
    Activations Density 0.637%

    No Known Activations