INDEX

Explanations

information related to individuals and their actions or controversies surrounding them

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their associated controversies or criticisms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

kefeller

-0.76

daq

-0.76

 ILCS

-0.76

FIG

-0.73

agues

-0.73

Num

-0.72

 GOODMAN

-0.72

TAIN

-0.72

Downloadha

-0.69

ĺħ

-0.68

POSITIVE LOGITS

 refusal

1.43

 handling

1.30

 unwillingness

1.23

 insistence

1.21

 inability

1.19

 portrayal

1.15

 involvement

1.14

 stance

1.14

 inaction

1.14

 dealings

1.13

Activations Density 0.205%

information related to individuals and their actions or controversies surrounding them

references to individuals and their associated controversies or criticisms

No Comments

No Known Activations