© 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. 8-RES-JB
    11. 4044
    Prev
    Next
    INDEX
    Explanations

    pronouns followed by verbs indicating actions

    oai_token-act-pair · gpt-3.5-turbo

    references to male and female subjects and their actions or states

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ©¶æ¥µ
    -0.75
    interstitial
    -0.67
    marked
    -0.65
    quartered
    -0.63
    enhagen
    -0.63
    rosse
    -0.63
     ILCS
    -0.60
    Sharp
    -0.60
     Electrical
    -0.59
    folios
    -0.59
    POSITIVE LOGITS
     succeed
    1.04
     succeeds
    1.03
     weren
    0.97
     fails
    0.93
    're
    0.89
     hadn
    0.89
     survives
    0.87
     chooses
    0.84
     didnt
    0.83
     ever
    0.82
    Activations Density 0.165%

    No Known Activations