INDEX

Explanations

phrases related to worms

oai_token-act-pair · gpt-3.5-turbo

references to worms 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 @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ctic

-0.79

ioch

-0.69

owered

-0.69

yles

-0.69

culated

-0.68

++++++++++++++++

-0.68

ulton

-0.68

rylic

-0.67

-----------

-0.66

trl

-0.64

POSITIVE LOGITS

hole

1.43

holes

1.24

worm

1.22

worms

1.19

 worms

1.05

 Worm

0.97

tail

0.97

 worm

0.97

fish

0.95

weed

0.89

Activations Density 0.012%

phrases related to worms

references to worms in various contexts

No Comments

No Known Activations