INDEX

Explanations

the word "these" followed by a description or object

oai_token-act-pair · gpt-3.5-turbo

references to previously mentioned concepts or items in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ĸļ

-0.91

abus

-0.85

osate

-0.80

opus

-0.76

atown

-0.74

slaught

-0.73

oru

-0.73

nect

-0.71

izont

-0.71

ntax

-0.69

POSITIVE LOGITS

 factors

1.19

 endeavors

0.99

 phenomena

0.94

 facets

0.92

 things

0.92

 aspects

0.91

 guys

0.89

 scenarios

0.89

 propositions

0.87

 situations

0.84

Activations Density 0.061%

the word "these" followed by a description or object

references to previously mentioned concepts or items in the text

No Comments

No Known Activations