INDEX

Explanations

keywords related to ranking, position, or hierarchy

oai_token-act-pair · gpt-3.5-turbo

references to rankings or positions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Goodbye

-0.69

fman

-0.66

drm

-0.62

stud

-0.62

amera

-0.62

ufact

-0.61

ctions

-0.61

 Wass

-0.61

vous

-0.61

xit

-0.60

POSITIVE LOGITS

 rank

4.00

 Rank

2.45

rank

2.36

 ranks

2.36

Rank

2.26

 ranking

2.16

 ranked

1.86

 Ranking

1.79

 Ranked

1.52

 rankings

1.50

Activations Density 0.011%

keywords related to ranking, position, or hierarchy

references to rankings or positions

No Comments

No Known Activations