© 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. 4-RES-JB
    11. 10376
    Prev
    Next
    INDEX
    Explanations

    phrases related to interviews or press conferences

    oai_token-act-pair · gpt-3.5-turbo

    references to news programming or discussions surrounding political events

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

    No Comments

    Negative Logits
     sake
    -0.74
    opsis
    -0.66
    ĨĴ
    -0.64
    emp
    -0.62
    hood
    -0.61
     arithmetic
    -0.60
     mol
    -0.60
    ength
    -0.59
     Mahjong
    -0.58
    steen
    -0.58
    POSITIVE LOGITS
    adr
    0.81
    orage
    0.75
    lyak
    0.74
    rolet
    0.70
    ritic
    0.70
    riot
    0.68
     Dull
    0.68
    OTOS
    0.67
    autions
    0.67
    iets
    0.66
    Activations Density 0.117%

    No Known Activations