© 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. MLP Post
    9. 2-MLP-SM
    11. 23805
    Prev
    Next
    INDEX
    Explanations

    negative sentiments or critiques

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/2-mlp-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_mlp_out
    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
    452
    +0.17
    1.0%
    326
    +0.13
    0.8%
    280
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    452
    +0.17
    0.02
    326
    +0.13
    0.02
    386
    +0.13
    0.01
    Negative Logits
    ographically
    -1.73
    ptions
    -1.59
    iled
    -1.53
     dedicated
    -1.42
    eting
    -1.42
     journey
    -1.35
     live
    -1.34
    assed
    -1.33
    ÂĹ
    -1.33
     rooms
    -1.32
    POSITIVE LOGITS
    ½
    4.08
    Ń
    4.05
    ł
    3.73
    §
    3.72
    ļ
    3.70
    °
    3.65
    Ĵ
    3.58
    ŀ
    3.55
    ↵  Âł
    3.52
    č↵č↵
    3.52
    Activations Density 0.223%

    No Known Activations