INDEX

Explanations

words related to political figures, particularly "Reagan."

oai_token-act-pair · gpt-3.5-turbo

references to Ronald Reagan and his administration

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enance

-0.70

hemat

-0.70

ĨĴ

-0.68

tered

-0.68

âĸ¬âĸ¬

-0.67

 Loaded

-0.65

razil

-0.63

âĸ¬

-0.62

ournals

-0.61

 Malays

-0.60

POSITIVE LOGITS

omics

1.29

 Reagan

0.95

ites

0.85

ism

0.83

 appoint

0.78

eland

0.78

ite

0.77

itus

0.75

ettes

0.74

isms

0.72

Activations Density 0.025%

words related to political figures, particularly "Reagan."

references to Ronald Reagan and his administration

No Comments

No Known Activations