INDEX

Explanations

the word "pipe"

oai_token-act-pair · gpt-3.5-turbo

references to "pipe" 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

ERAL

-0.80

nesota

-0.80

enance

-0.80

é¾įå

-0.78

neapolis

-0.75

ean

-0.75

ãħĭ

-0.74

 GOODMAN

-0.73

eanor

-0.70

orian

-0.70

POSITIVE LOGITS

pipe

1.18

 pipe

1.11

 pipes

1.01

 Pipe

0.95

 piping

0.91

 cleaners

0.90

 cleaner

0.84

tob

0.82

cone

0.77

 smoking

0.76

Activations Density 0.011%

the word "pipe"

references to "pipe" in various contexts

No Comments

No Known Activations