© 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. 0-RES-JB
    11. 1284
    Prev
    Next
    INDEX
    Explanations

    expressions of enthusiasm or emphasis

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.0.hook_resid_post_16384_anthropic
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    HuggingFaceFW/fineweb
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    528
    +0.10
    0.3%
    1335
    +0.08
    0.3%
    517
    +0.07
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    972
    +0.10
    0.03
    963
    +0.08
    0.03
    1150
    +0.07
    0.01
    Negative Logits
     continue
    -0.65
     realize
    -0.61
     overcome
    -0.60
     put
    -0.59
     meet
    -0.57
     will
    -0.57
     may
    -0.56
     understand
    -0.56
     know
    -0.56
    要
    -0.56
    POSITIVE LOGITS
     viciss
    1.36
     estimat
    1.31
     velours
    1.29
     kram
    1.27
     Græ
    1.26
     Somal
    1.26
     wien
    1.26
     franz
    1.26
     immen
    1.25
     accla
    1.25
    Activations Density 0.171%

    No Known Activations