© 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. 10-RES-JB
    11. 18540
    Prev
    Next
    INDEX
    Explanations

    instances of the phrase "do what he/she/they has/have/had to do"

    oai_token-act-pair · gpt-3.5-turbo

    phrases that express the idea of performing actions or fulfilling roles

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    arer
    -0.69
    »Ĵ
    -0.65
    ounter
    -0.65
     circulation
    -0.64
     gulf
    -0.63
     reconcil
    -0.63
    ;;;;;;;;;;;;
    -0.63
     identification
    -0.62
    etheless
    -0.58
     rupture
    -0.58
    POSITIVE LOGITS
     preach
    0.79
    done
    0.77
     preached
    0.76
     gotta
    0.75
     commanded
    0.74
     instructed
    0.74
     commandments
    0.73
    erd
    0.71
     undertook
    0.70
     asked
    0.70
    Activations Density 0.106%

    No Known Activations