© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 2-RES-JB
    11. 11155
    Prev
    Next
    INDEX
    Explanations

    words related to intense physical actions or states of being

    oai_token-act-pair · gpt-3.5-turbo

    terms associated with specific characteristics or conditions that reflect intensity or severity

    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
    GV
    -0.43
    NL
    -0.42
    orem
    -0.42
     Lov
    -0.41
    equ
    -0.41
    ERO
    -0.41
    Hor
    -0.40
    abus
    -0.40
    Buy
    -0.39
    BIT
    -0.39
    POSITIVE LOGITS
    etheless
    0.57
    minist
    0.50
    terness
    0.50
    Redditor
    0.47
    !--
    0.43
    NetMessage
    0.43
     Pradesh
    0.42
    gery
    0.42
     advis
    0.42
     dexter
    0.41
    Activations Density 1.005%

    No Known Activations