INDEX

Explanations

words related to negative effects or potential harm

oai_token-act-pair · gpt-3.5-turbo

terms related to damage or harm to various entities or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

DragonMagazine

-0.76

az

-0.72

mad

-0.71

 photographed

-0.65

ãĤ©

-0.64

 Edited

-0.63

lich

-0.62

BuyableInstoreAndOnline

-0.60

Mut

-0.60

lov

-0.59

POSITIVE LOGITS

 havoc

1.22

 morale

1.17

 livelihood

0.94

 credibility

0.90

 immersion

0.89

 perceptions

0.84

 productivity

0.84

 competitiveness

0.82

 efficiency

0.76

 mobility

0.74

Activations Density 0.216%

words related to negative effects or potential harm

terms related to damage or harm to various entities or situations

No Comments

No Known Activations