INDEX

Explanations

mentions of buses

oai_token-act-pair · gpt-3.5-turbo

references to buses

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

sson

-0.75

 marrow

-0.75

 Martial

-0.73

aneous

-0.69

arial

-0.69

vironment

-0.64

 Christensen

-0.63

 Centauri

-0.62

 forth

-0.60

 Citizenship

-0.60

POSITIVE LOGITS

loads

1.21

load

1.07

hel

0.98

boys

0.96

kers

0.86

INESS

0.85

boy

0.84

hy

0.82

dump

0.79

driver

0.77

Activations Density 0.019%

mentions of buses

references to buses

No Comments

No Known Activations