INDEX

Explanations

phrases indicating the end or completion of something

oai_token-act-pair · gpt-3.5-turbo

phrases associated with summarizing or concluding statements

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

loc

-0.74

usercontent

-0.69

 monkeys

-0.67

finder

-0.65

ied

-0.64

 CrossRef

-0.63

 guid

-0.61

iru

-0.60

 volunteers

-0.58

 memos

-0.57

POSITIVE LOGITS

OX

0.74

 festivities

0.70

IF

0.69

SHIP

0.68

OWS

0.67

nings

0.66

loading

0.65

 EDITION

0.63

APH

0.62

SX

0.61

Activations Density 0.180%

phrases indicating the end or completion of something

phrases associated with summarizing or concluding statements

No Comments

No Known Activations