INDEX

Explanations

words related to decision-making and opinions

oai_token-act-pair · gpt-3.5-turbo

phrases related to clarity and understanding in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Quit

-0.54

ggles

-0.48

 Sierra

-0.47

 Continued

-0.46

 former

-0.46

cknowled

-0.45

 assisted

-0.44

 formerly

-0.44

 Selected

-0.43

 Recover

-0.42

POSITIVE LOGITS

 toget

0.62

edIn

0.61

 thous

0.61

daq

0.58

 vulner

0.57

aeda

0.56

omorphic

0.54

 Seym

0.54

agra

0.51

wcs

0.51

Activations Density 6.585%

words related to decision-making and opinions

phrases related to clarity and understanding in various contexts

No Comments

No Known Activations