© 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. 3-RES-JB
    11. 15750
    Prev
    Next
    INDEX
    Explanations

    terms related to the abortion debate, specifically terms related to the pro-life position

    oai_token-act-pair · gpt-3.5-turbo

    terms related to pro-choice and pro-life positions in the context of abortion

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    âĸ¬
    -0.95
    ahime
    -0.80
    RAW
    -0.74
     Obj
    -0.73
    ripp
    -0.68
     Objective
    -0.68
     Username
    -0.67
     Estimated
    -0.66
     Passive
    -0.65
    æ°
    -0.65
    POSITIVE LOGITS
    rights
    1.01
    iferation
    0.88
     rights
    0.86
    ervative
    0.84
    ervatives
    0.84
    ertodd
    0.83
     advocates
    0.82
    ansas
    0.80
    choice
    0.80
     Parenthood
    0.80
    Activations Density 0.025%

    No Known Activations