© 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. 2-RES-JB
    6. 18585
    Prev
    Next
    INDEX
    Explanations

    phrases related to acts of sexual abuse

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the root word "molest" and its variations in various contexts, primarily relating to the act of molestation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 2-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    2
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Defenders
    -0.75
    DragonMagazine
    -0.70
    winner
    -0.69
    ĪĴ
    -0.69
    hani
    -0.68
    Condition
    -0.63
    Winner
    -0.62
     Founders
    -0.62
     Chosen
    -0.62
     Millennium
    -0.62
    POSITIVE LOGITS
     mol
    1.02
    estation
    0.96
    mol
    0.94
    riages
    0.93
    reated
    0.87
    asses
    0.85
    osexual
    0.85
    esting
    0.83
    oths
    0.83
    eties
    0.81
    Activations Density 0.005%

    No Known Activations