© 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 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. 10868
    Prev
    Next
    INDEX
    Explanations

    expressions of contrast or opposition

    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
    478
    +0.19
    0.6%
    381
    +0.12
    0.4%
    1937
    +0.11
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    478
    +0.19
    0.08
    658
    +0.12
    0.09
    1937
    +0.11
    0.08
    Negative Logits
     kafe
    -1.26
     silikon
    -1.22
     keramik
    -1.18
     optik
    -1.17
     kön
    -1.17
     seksi
    -1.17
     antik
    -1.16
     maksi
    -1.14
     protokol
    -1.13
     kosme
    -1.04
    POSITIVE LOGITS
     intersper
    1.04
     McLaugh
    0.87
     ecru
    0.85
     hairc
    0.82
     peppa
    0.80
     unspeak
    0.80
     cushi
    0.80
     apprehen
    0.80
     McInt
    0.80
     disreg
    0.79
    Activations Density 0.432%

    No Known Activations