INDEX

Explanations

lists of entities separated by commas

oai_token-act-pair · gpt-3.5-turbo

references to countries and related statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

herent

-0.72

 metic

-0.65

usional

-0.65

 proble

-0.64

vered

-0.62

natureconservancy

-0.61

iple

-0.60

estone

-0.59

grain

-0.59

orem

-0.58

POSITIVE LOGITS

etc

1.09

 alas

0.85

 albeit

0.84

 although

0.82

huh

0.79

 Kinnikuman

0.77

aka

0.77

 whereas

0.74

 preferably

0.72

0.71

Activations Density 1.303%

lists of entities separated by commas

references to countries and related statistics

No Comments

No Known Activations