INDEX

Explanations

adverbs and intensifiers

oai_token-act-pair · gpt-3.5-turbo

phrases indicative of negation or denial

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

catentry

-0.75

LIN

-0.65

 Almighty

-0.62

 Naples

-0.59

Net

-0.58

study

-0.58

 Palest

-0.56

MAC

-0.54

 Morse

-0.54

Mos

-0.54

POSITIVE LOGITS

 anymore

0.98

 bothered

0.85

 bother

0.74

 anywhere

0.72

ANY

0.71

ymes

0.71

 necessarily

0.70

 Enough

0.69

*/(

0.68

any

0.66

Activations Density 0.271%

adverbs and intensifiers

phrases indicative of negation or denial

No Comments

No Known Activations