INDEX

Explanations

examples that follow the 'for example' format

oai_token-act-pair · gpt-3.5-turbo

the letter 'g' in various contexts

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

rawdownloadcloneembedreportprint

-0.64

Rated

-0.62

 regist

-0.60

 Wonderful

-0.59

FINE

-0.57

 heats

-0.57

 intern

-0.56

rontal

-0.56

 completes

-0.56

DEBUG

-0.55

POSITIVE LOGITS

asp

0.90

raphics

0.87

.,

0.79

ardless

0.75

audi

0.75

raham

0.74

mma

0.73

ascript

0.73

oths

0.72

ossip

0.70

Activations Density 0.010%

examples that follow the 'for example' format

the letter 'g' in various contexts

No Comments

No Known Activations