© 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. Joseph Bloom · Initial Gemma-2B SAEs
    5. Gemma-2B
    7. Residual Stream
    9. 10-RES-JB
    11. 4235
    Prev
    Next
    INDEX
    Explanations

    phrases that describe methods or ways to approach a subject

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.10.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    chanind/openwebtext-gemma
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.10.hook_resid_post
    Hook Layer
    10
    Architecture
    standard
    Context Size
    1,024
    Dataset
    ctigges/openwebtext-gemma-1024-cl
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    50
    +0.39
    1.6%
    604
    +0.08
    0.3%
    537
    +0.08
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    16
    +0.39
    0.08
    80
    +0.08
    0.04
    1703
    +0.08
    0.06
    Negative Logits
    <bos>
    -3.15
    ⓧ
    -0.81
    <?
    -0.77
    /**
    -0.75
     intersper
    -0.67
    /***
    -0.62
    engeance
    -0.60
     inaugurate
    -0.60
     endow
    -0.59
     disbur
    -0.59
    POSITIVE LOGITS
     habet
    1.02
     Minang
    1.01
     cæ
    0.99
     eorum
    0.97
     ejus
    0.94
     quæ
    0.93
     potest
    0.91
     tibi
    0.86
     quoque
    0.86
     omnes
    0.85
    Activations Density 0.828%

    No Known Activations