INDEX

Explanations

races or events

oai_token-act-pair · gpt-3.5-turbo

occurrences of "Grand Prix" and related racing terms

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

itable

-0.78

oks

-0.75

izard

-0.74

ded

-0.70

gren

-0.70

amus

-0.70

itably

-0.70

bread

-0.69

stocks

-0.67

Else

-0.67

POSITIVE LOGITS

 Prix

1.25

 Weekend

0.84

 Racer

0.79

 Showdown

0.77

 racing

0.76

 Tournament

0.76

 Arena

0.75

Rac

0.75

FactoryReloaded

0.74

 Championships

0.74

Activations Density 0.011%

races or events

occurrences of "Grand Prix" and related racing terms

No Comments

No Known Activations