INDEX

Explanations

words related to denial or refusal

oai_token-act-pair · gpt-3.5-turbo

instances of the word "deny" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ellow

-0.74

utenberg

-0.74

ram

-0.73

OPER

-0.72

ICAN

-0.71

orough

-0.71

gm

-0.71

ROM

-0.71

prototype

-0.71

amen

-0.70

POSITIVE LOGITS

 deny

1.26

 denies

1.26

 denying

1.10

 denied

1.02

 denial

0.95

 outright

0.87

 excuses

0.81

 rehe

0.77

 dismiss

0.73

 discriminate

0.72

Activations Density 0.007%

words related to denial or refusal

instances of the word "deny" in various contexts

No Comments

No Known Activations