INDEX

Explanations

phrases indicating additional information or further reading

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "Read more."

oai_token-act-pair · gpt-4o-mini Triggered 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

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ļé

-0.76

XL

-0.67

ISO

-0.64

xtap

-0.63

itialized

-0.61

©¶æ¥µ

-0.61

wagen

-0.59

pect

-0.58

arsity

-0.58

eros

-0.58

POSITIVE LOGITS

 Â»

0.71

 than

0.66

 âĨĴ

0.65

 âĢº

0.56

 Juliet

0.54

0.52

 VIDEOS

0.51

maker

0.51

 auth

0.51

>>

0.50

Activations Density 0.025%

phrases indicating additional information or further reading

instances of the phrase "Read more."

No Comments

No Known Activations