INDEX

Explanations

questions ending with a question mark

oai_token-act-pair · gpt-3.5-turbo

questions marked by a question mark

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

aper

-0.82

reed

-0.72

 exiled

-0.72

 manif

-0.72

apers

-0.69

onto

-0.69

athe

-0.67

 worm

-0.67

 drowned

-0.66

ems

-0.66

POSITIVE LOGITS

Â¶

0.99

.?

0.96

 Check

0.95

 Nope

0.95

 Answer

0.93

 Does

0.92

 Where

0.91

 Probably

0.88

 Availability

0.87

?:

0.87

Activations Density 0.095%

questions ending with a question mark

questions marked by a question mark

No Comments

No Known Activations