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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 881
    Prev
    Next
    INDEX
    Explanations

    specific time references such as "Morning" or "Thursday."

    oai_token-act-pair · gpt-3.5-turboTriggered 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
    1023
    +0.13
    0.4%
    1757
    +0.11
    0.4%
    1805
    +0.11
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1805
    +0.13
    0.03
    1023
    +0.11
    0.03
    655
    +0.11
    0.03
    Negative Logits
     rendono
    -0.56
     SqlDbType
    -0.55
    hamdu
    -0.49
     possano
    -0.49
     itong
    -0.48
     quí
    -0.48
     Ottobre
    -0.47
     sembrano
    -0.47
     numerus
    -0.46
     abbiano
    -0.46
    POSITIVE LOGITS
     morning
    1.36
    morning
    1.32
    Morning
    1.23
     Morning
    1.22
     mornings
    1.19
     MORNING
    1.18
     breakfast
    0.94
     Breakfast
    0.84
     Morn
    0.83
     breakfasts
    0.81
    Activations Density 0.088%

    No Known Activations