© 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. 11-RES-JB
    11. 9125
    Prev
    Next
    INDEX
    Explanations

    words related to defamation or negative accusations

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "defame" and its variations, indicating a focus on accusations or slander

    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
    sth
    -0.68
     roundup
    -0.63
     fitt
    -0.60
     slic
    -0.60
    åŃ
    -0.58
     terminal
    -0.58
    BIL
    -0.58
     attachments
    -0.58
     twists
    -0.58
    QB
    -0.57
    POSITIVE LOGITS
    erence
    1.46
    ection
    1.42
    ector
    1.41
    ensible
    1.39
    erential
    1.37
    lection
    1.36
    lected
    1.36
    ected
    1.35
    acement
    1.33
    lated
    1.32
    Activations Density 0.013%

    No Known Activations