INDEX

Explanations

proper nouns known for specific attributes or characteristics

oai_token-act-pair · gpt-3.5-turbo

phrases indicating something is known for a particular quality or attribute

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

Reviewed

-0.72

ogether

-0.69

ieties

-0.67

VR

-0.66

FP

-0.65

Balt

-0.64

urs

-0.62

III

-0.61

down

-0.61

eros

-0.58

POSITIVE LOGITS

bidden

0.93

geries

0.93

gery

0.85

 daring

0.82

ked

0.75

Ĥª

0.75

 awhile

0.70

 example

0.70

 decades

0.68

 storing

0.68

Activations Density 0.066%

proper nouns known for specific attributes or characteristics

phrases indicating something is known for a particular quality or attribute

No Comments

No Known Activations