© 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
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 11-RES-JB
    6. 894
    Prev
    Next
    INDEX
    Explanations

    instances where something is being done, whether it's in the context of various activities or precautions taken

    oai_token-act-pair · gpt-3.5-turbo

    the conditional phrase "whether it" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Reviewer
    -0.70
    į
    -0.66
    IFT
    -0.66
    GROUND
    -0.61
     newcom
    -0.59
    å½
    -0.58
     Genius
    -0.58
    package
    -0.58
    former
    -0.57
    OME
    -0.56
    POSITIVE LOGITS
     involves
    1.36
     hurts
    1.17
     coincides
    1.15
     rains
    1.15
     occurs
    1.13
     means
    1.13
     affects
    1.13
     happens
    1.13
    's
    1.12
     entails
    1.12
    Activations Density 0.135%

    No Known Activations