INDEX

Explanations

phrases that summarize information or opinions

oai_token-act-pair · gpt-3.5-turbo

phrases related to summarization and encapsulation of information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ellen

-0.75

eworld

-0.72

eland

-0.72

 tradem

-0.71

hips

-0.67

ali

-0.67

ship

-0.66

edy

-0.64

entity

-0.63

duct

-0.63

POSITIVE LOGITS

 summar

1.09

 summed

1.09

 succinct

0.96

ãĤ©

0.85

 summarized

0.83

 encaps

0.81

 summarize

0.80

STAT

0.79

 summarizes

0.77

 concise

0.76

Activations Density 0.023%

phrases that summarize information or opinions

phrases related to summarization and encapsulation of information

No Comments

No Known Activations