INDEX

Explanations

phrases related to past employment and job roles

oai_token-act-pair · gpt-3.5-turbo

phrases describing occupations or roles

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

oche

-0.75

yrs

-0.74

redits

-0.70

tons

-0.68

/$

-0.66

rous

-0.65

rogens

-0.62

ALS

-0.62

raq

-0.62

ru

-0.62

POSITIVE LOGITS

pires

1.19

pired

1.09

an

1.00

0.94

 deputy

0.91

 ambassador

0.90

 curator

0.88

 assistant

0.86

 director

0.86

 translator

0.86

Activations Density 0.089%

phrases related to past employment and job roles

phrases describing occupations or roles

No Comments

No Known Activations