INDEX

Explanations

phrases related to tutorials or mentoring experiences

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of the word "tut" and its variants, indicating references to tutorials

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 saline

-0.82

 captivity

-0.75

å£«

-0.68

 mine

-0.65

ISSION

-0.65

 Lima

-0.64

Unknown

-0.63

 Franch

-0.63

theless

-0.61

oples

-0.61

POSITIVE LOGITS

tle

1.23

tut

1.14

icket

1.03

te

1.01

ickets

0.97

ypes

0.96

anium

0.95

Tut

0.95

ilities

0.94

tops

0.93

Activations Density 0.025%

phrases related to tutorials or mentoring experiences

repeated mentions of the word "tut" and its variants, indicating references to tutorials

No Comments

No Known Activations