INDEX

Explanations

specific locations or organizations mentioned in news articles

oai_token-act-pair · gpt-3.5-turbo

references to image captions or descriptors

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Kinder

-0.58

Tid

-0.57

 Corps

-0.55

gut

-0.55

rous

-0.53

 produ

-0.52

 whistle

-0.52

plain

-0.51

 todd

-0.50

 Overse

-0.50

POSITIVE LOGITS

0.74

PHOTOS

0.74

"/

0.71

 guiIcon

0.70

=/

0.68

":"/

0.68

interstitial

0.64

 Images

0.63

icons

0.63

 Courtesy

0.62

Activations Density 0.010%

specific locations or organizations mentioned in news articles

references to image captions or descriptors

No Comments

No Known Activations