INDEX

Explanations

phrases indicating additional information or details

oai_token-act-pair · gpt-3.5-turbo

references to updates and additional information

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

isa

-0.65

gall

-0.65

soever

-0.64

 Murd

-0.61

pires

-0.61

imaru

-0.60

conn

-0.59

 Steal

-0.58

 inval

-0.58

 testified

-0.58

POSITIVE LOGITS

 sake

0.86

 details

0.84

 explanation

0.83

 specifics

0.77

ummies

0.75

.):

0.73

 purposes

0.73

 clarification

0.72

 inspiration

0.71

 reasons

0.70

Activations Density 0.239%

phrases indicating additional information or details

references to updates and additional information

No Comments

No Known Activations