© 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. Joseph Bloom · Initial Gemma-2B-IT SAEs
    5. Gemma-2B-IT
    7. Residual Stream
    9. 12-RES-JB
    11. 3930
    Prev
    Next
    INDEX
    Explanations

    proper nouns, especially names of places and people

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    The common pattern involves the use of abbreviations with periods, typically denoting a person's name or a legal citation, alongside a mix of formal and conversational language within diverse contexts.

    eleuther_acts_top20 · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    standard
    Context Size
    1,024
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    227
    +0.11
    0.3%
    1905
    +0.11
    0.3%
    876
    +0.09
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    981
    +0.11
    0.06
    1471
    +0.11
    0.05
    2010
    +0.09
    0.04
    Negative Logits
    ManagedBean
    -0.95
    astéro
    -0.95
    ngdoc
    -0.95
     Exacts
    -0.94
     Walkover
    -0.93
    Galería
    -0.92
     يتيمه
    -0.91
    Premios
    -0.91
    Argumento
    -0.91
    Εκ
    -0.91
    POSITIVE LOGITS
     intersper
    1.78
     encomp
    1.69
     maneu
    1.69
     impra
    1.68
     increa
    1.66
     shenan
    1.65
     attemp
    1.61
     fuf
    1.60
     milf
    1.58
     reluct
    1.57
    Activations Density 0.202%

    No Known Activations