INDEX

Explanations

proper nouns or names

oai_token-act-pair · gpt-3.5-turbo

proper nouns associated with locations and individuals noted in photographs

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

 dracon

-0.73

heartedly

-0.73

abiding

-0.70

luaj

-0.69

 ACTIONS

-0.67

 earnest

-0.65

ãĢİ

-0.61

 succeeding

-0.60

Â¶

-0.60

 totaling

-0.60

POSITIVE LOGITS

photos

0.94

obal

0.80

 Photos

0.80

File

0.78

asma

0.78

veyard

0.76

 Photography

0.74

images

0.72

Hau

0.72

 screenshot

0.72

Activations Density 0.136%

proper nouns or names

proper nouns associated with locations and individuals noted in photographs

No Comments

No Known Activations