© 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 · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 1560
    Prev
    Next
    INDEX
    Explanations

    words ending in 'r' with different activation values, where 'r' is activated the strongest

    oai_token-act-pair · gpt-3.5-turbo

    the letter "r" in various contexts throughout the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     moot
    -0.60
     Beckham
    -0.58
    izations
    -0.58
     yourselves
    -0.57
     countdown
    -0.56
     Cec
    -0.55
     commend
    -0.54
     furthe
    -0.53
     Tanaka
    -0.53
     Clockwork
    -0.52
    POSITIVE LOGITS
    r
    3.85
    rs
    2.31
    rn
    2.14
    rl
    2.05
    rr
    1.93
    rf
    1.86
    rg
    1.83
    rb
    1.83
    rab
    1.81
    rh
    1.79
    Activations Density 0.034%

    No Known Activations