INDEX

Explanations

prepositions followed by words related to conflict or controversy

oai_token-act-pair · gpt-3.5-turbo

recurring themes of conflict or disputes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

uum

-0.88

illary

-0.83

GH

-0.71

KE

-0.71

ï¸

-0.69

hesis

-0.69

OH

-0.67

////////////////////////////////

-0.66

uity

-0.66

 awaits

-0.65

POSITIVE LOGITS

tones

1.01

hang

0.92

lord

0.92

loading

0.85

reaching

0.84

haul

0.84

drive

0.82

population

0.80

priced

0.79

payment

0.78

Activations Density 0.073%

prepositions followed by words related to conflict or controversy

recurring themes of conflict or disputes

No Comments

No Known Activations