INDEX

Explanations

years mentioned between 1997 and 2007

oai_token-act-pair · gpt-3.5-turbo

references to specific years, particularly in the late 1990s

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hop

-0.78

hatt

-0.73

edo

-0.71

giving

-0.69

edIn

-0.68

EStreamFrame

-0.67

holders

-0.66

tro

-0.65

holder

-0.63

groups

-0.63

POSITIVE LOGITS

ãĤ¦ãĤ¹

0.79

ĸļ

0.77

-'

0.77

ãĥĥãĥī

0.69

ãĥŁ

0.69

 Reincarn

0.66

 Newsweek

0.65

ined

0.64

uncture

0.63

å¹

0.61

Activations Density 0.029%

years mentioned between 1997 and 2007

references to specific years, particularly in the late 1990s

No Comments

No Known Activations