INDEX

Explanations

words related to anger or strong negative emotions towards someone or something

oai_token-act-pair · gpt-3.5-turbo

expressions of anger or irritation

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

•layer 5 - anger

No Comments

Negative Logits

 livest

-0.93

age

-0.66

glas

-0.66

 Notting

-0.66

 faire

-0.65

 lucid

-0.65

fficiency

-0.62

ipeg

-0.61

 Refresh

-0.61

 liner

-0.61

POSITIVE LOGITS

terness

0.93

untled

0.90

 bitterly

0.83

ire

0.81

dy

0.76

 flared

0.76

actionDate

0.75

SHIP

0.75

 hostility

0.74

 angrily

0.74

Activations Density 0.116%

words related to anger or strong negative emotions towards someone or something

expressions of anger or irritation

No Comments

No Known Activations