INDEX

Explanations

references to squeezing or pressing of something

oai_token-act-pair · gpt-3.5-turbo

words related to "lee" or entities that begin with "le."

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

ãĥ¼ãĥĨãĤ£

-1.02

âĶģ

-0.94

ĸļ

-0.89

ãĥ¼ãĥĨ

-0.83

aneers

-0.81

ãģĨ

-0.76

ãĥ´ãĤ¡

-0.75

MODE

-0.75

etheless

-0.74

ĳå£«

-0.71

POSITIVE LOGITS

opard

1.06

isure

1.01

ather

0.94

icester

0.93

lla

0.93

aven

0.92

mons

0.91

lean

0.89

gging

0.88

yton

0.87

Activations Density 0.015%

references to squeezing or pressing of something

words related to "lee" or entities that begin with "le."

No Comments

No Known Activations