© 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. 23225
    Prev
    Next
    INDEX
    Explanations

    chemical elements, particularly iron (Fe)

    oai_token-act-pair · gpt-3.5-turbo

    references to the element Iron (Fe)

    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
     nav
    -0.85
    vag
    -0.75
    ascade
    -0.70
    tor
    -0.69
    lock
    -0.69
     Kenya
    -0.68
     batters
    -0.68
     pod
    -0.67
    ARC
    -0.66
     NAV
    -0.66
    POSITIVE LOGITS
     Fe
    3.91
    Fe
    2.56
     fe
    1.44
     FE
    1.39
    fe
    1.20
     Fein
    1.19
     Ca
    1.13
    Fa
    1.12
     Feldman
    1.11
     Feinstein
    1.11
    Activations Density 0.024%

    No Known Activations