INDEX

Explanations

words related to negative political events or actions

oai_token-act-pair · gpt-3.5-turbo

negative prefixes or terms associated with disabilities and social issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 breed

-0.74

 Nurse

-0.68

 offence

-0.67

 answ

-0.67

 defeating

-0.65

 quantity

-0.65

 Lord

-0.65

 practice

-0.65

 DVDs

-0.65

 seiz

-0.63

POSITIVE LOGITS

tur

1.27

nuclear

1.21

direction

1.20

technical

1.18

cultural

1.17

electric

1.15

trop

1.13

economic

1.13

social

1.12

norm

1.10

Activations Density 0.051%

words related to negative political events or actions

negative prefixes or terms associated with disabilities and social issues

No Comments

No Known Activations