INDEX

Explanations

phrases related to technical or computational topics

oai_token-act-pair · gpt-3.5-turbo

phrases related to various types of measurements and metrics in different contexts

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

 overtime

-0.70

 agre

-0.68

 toile

-0.67

 neighb

-0.66

wet

-0.66

wip

-0.66

 retard

-0.65

 destro

-0.63

 obscene

-0.62

 star

-0.61

POSITIVE LOGITS

ccording

1.14

////////////////////////////////

0.98

Advertisements

0.97

Whether

0.97

Spoiler

0.97

Asset

0.97

Since

0.95

========

0.95

Given

0.94

There

0.94

Activations Density 0.232%

phrases related to technical or computational topics

phrases related to various types of measurements and metrics in different contexts

No Comments

No Known Activations