INDEX

Explanations

phrases related to unspecified concepts or ideas

oai_token-act-pair · gpt-3.5-turbo

references to unspecified or vague concepts

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

ortex

-0.76

rises

-0.70

oise

-0.68

de

-0.67

 Telecommunications

-0.66

ilt

-0.66

DOS

-0.64

oard

-0.64

éĹĺ

-0.62

design

-0.62

POSITIVE LOGITS

 else

1.41

Else

1.22

 akin

0.96

 shameful

0.83

 Else

0.81

 intangible

0.80

 worthwhile

0.79

 resembling

0.78

 unusual

0.77

 unheard

0.77

Activations Density 0.037%

phrases related to unspecified concepts or ideas

references to unspecified or vague concepts

No Comments

No Known Activations