INDEX

Explanations

proper nouns or titles typically featured in news articles

oai_token-act-pair · gpt-3.5-turbo

references to significant events or crises

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

oise

-0.81

tons

-0.64

opsis

-0.64

 manufact

-0.63

osaurus

-0.63

 Fine

-0.63

ilo

-0.62

 Doodle

-0.62

 Pony

-0.62

ean

-0.61

POSITIVE LOGITS

ccording

0.80

BBC

0.71

jer

0.67

accompan

0.66

Bus

0.66

Laura

0.66

dan

0.65

Posted

0.64

inav

0.64

 bishops

0.62

Activations Density 0.129%

proper nouns or titles typically featured in news articles

references to significant events or crises

No Comments

No Known Activations