INDEX

Explanations

phrases indicating logical or reasonable explanations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating comprehension or logical reasoning

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

quart

-0.83

psy

-0.78

aunting

-0.73

ching

-0.72

enne

-0.71

 helicop

-0.70

leted

-0.67

rave

-0.66

venge

-0.66

rising

-0.65

POSITIVE LOGITS

edIn

0.74

ãħĭ

0.73

rawdownloadcloneembedreportprint

0.72

 ..............

0.71

 Tradable

0.70

 partName

0.68

INESS

0.68

GROUP

0.68

SourceFile

0.68

 Marketable

0.68

Activations Density 0.024%

phrases indicating logical or reasonable explanations

phrases indicating comprehension or logical reasoning

No Comments

No Known Activations