INDEX

Explanations

phrases expressing disapproval, critique, or criticism

oai_token-act-pair · gpt-3.5-turbo

phrases expressing strong disapproval or criticism

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

thood

-0.77

aido

-0.75

enhagen

-0.74

enne

-0.74

toggle

-0.68

vern

-0.68

metadata

-0.67

 mirac

-0.66

ipl

-0.66

erential

-0.65

POSITIVE LOGITS

 unacceptable

1.43

 disgrace

1.35

 violates

1.33

 unworthy

1.29

 undermines

1.27

 disrespectful

1.26

 endanger

1.25

 detrimental

1.23

 jeopard

1.22

 harming

1.22

Activations Density 0.446%

phrases expressing disapproval, critique, or criticism

phrases expressing strong disapproval or criticism

No Comments

No Known Activations