INDEX

Explanations

threatening or confrontational language

oai_token-act-pair · gpt-3.5-turbo

aggressive language and threats

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 preceding

-0.75

quartered

-0.75

 lacked

-0.70

 opting

-0.69

 unusually

-0.69

imilar

-0.69

pired

-0.68

 strikingly

-0.68

 seemingly

-0.68

 overshadowed

-0.66

POSITIVE LOGITS

 tomorrow

1.20

morrow

1.06

 whoever

1.02

 someday

1.00

!"

0.98

 ..."

0.96

 ASAP

0.95

 fuckin

0.93

 âĢ¦"

0.92

 ______

0.88

Activations Density 0.718%

threatening or confrontational language

aggressive language and threats

No Comments

No Known Activations