INDEX

Explanations

words related to disorder or chaos

oai_token-act-pair · gpt-3.5-turbo

references to disorder or chaos

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

cember

-0.83

undai

-0.78

vation

-0.72

 livest

-0.71

©¶æ

-0.68

rendered

-0.68

CVE

-0.66

cki

-0.65

aldi

-0.65

 obser

-0.63

POSITIVE LOGITS

engers

1.48

iah

1.13

enger

0.98

es

0.94

aging

0.82

 havoc

0.81

cat

0.80

aged

0.79

romeda

0.79

mess

0.78

Activations Density 0.023%

words related to disorder or chaos

references to disorder or chaos

No Comments

No Known Activations