© 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. 8-RES-JB
    11. 7995
    Prev
    Next
    INDEX
    Explanations

    words related to profanity or inappropriate language

    oai_token-act-pair · gpt-3.5-turbo

    words related to profanity or explicit language

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    warm
    -0.84
    MENTS
    -0.81
    forth
    -0.73
    doors
    -0.73
     halfway
    -0.68
    MENT
    -0.65
    nuts
    -0.62
    wolves
    -0.62
     Nou
    -0.61
    boat
    -0.61
    POSITIVE LOGITS
    essor
    1.36
    iles
    1.33
    iler
    1.28
    iciency
    1.27
    ilers
    1.25
    essors
    1.22
    iled
    1.19
    ession
    1.14
    icient
    1.14
    iling
    1.10
    Activations Density 0.026%

    No Known Activations