© 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. 10-RES-JB
    11. 10345
    Prev
    Next
    INDEX
    Explanations

    hyphenated phrases with "all" that usually have a negative connotation

    oai_token-act-pair · gpt-3.5-turbo

    phrases that indicate all-male groups or establishments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Caption
    -0.70
     Chains
    -0.65
     Reloaded
    -0.63
     Lumpur
    -0.63
     Stream
    -0.62
     sacrific
    -0.61
     RTX
    -0.59
     erosion
    -0.59
     Hilbert
    -0.58
     puberty
    -0.58
    POSITIVE LOGITS
    purpose
    1.32
    important
    1.22
    consuming
    1.21
    too
    1.17
    around
    1.16
    powerful
    1.16
    female
    1.15
    caps
    1.15
    male
    1.14
    star
    1.11
    Activations Density 0.027%

    No Known Activations