© 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. 0-RES-JB
    11. 19945
    Prev
    Next
    INDEX
    Explanations

    phrases related to the name "Hook"

    oai_token-act-pair · gpt-3.5-turbo

    mentions of the term "Hook" and related references to the Sandy Hook tragedy

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     parity
    -0.82
     dil
    -0.76
     glac
    -0.70
    UE
    -0.70
     separat
    -0.68
     ultr
    -0.68
    acia
    -0.67
     sweating
    -0.66
     diversity
    -0.65
     radi
    -0.65
    POSITIVE LOGITS
     Hook
    4.05
    hook
    1.95
     Newtown
    1.25
     Hood
    1.19
     hooks
    1.18
     Winc
    1.10
     Hog
    1.09
     Screw
    1.06
     hook
    1.04
     Tackle
    1.02
    Activations Density 0.017%

    No Known Activations