INDEX

Explanations

captions in images

oai_token-act-pair · gpt-3.5-turbo

mentions of image captions and their corresponding attributes

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

<|endoftext|>

-0.74

aturdays

-0.69

akespeare

-0.68

onest

-0.67

bom

-0.61

 recl

-0.61

 territ

-0.60

ury

-0.60

stab

-0.59

 reborn

-0.59

POSITIVE LOGITS

 GOODMAN

0.92

 Gladiator

0.75

Phys

0.75

UTERS

0.73

 Javascript

0.68

 IMAGES

0.65

 Mandatory

0.65

 Immun

0.64

Ability

0.63

itars

0.63

Activations Density 0.094%

captions in images

mentions of image captions and their corresponding attributes

No Comments

No Known Activations