INDEX

Explanations

instances of the word "to" followed by a verb

oai_token-act-pair · gpt-3.5-turbo

instances of the word "to" indicating directional action or intentions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

selves

-0.90

enance

-0.66

 entit

-0.64

nav

-0.64

fortunately

-0.64

terday

-0.63

eanor

-0.63

angered

-0.63

heast

-0.63

ankind

-0.61

POSITIVE LOGITS

bed

0.99

pless

0.96

 extremes

0.95

 lengths

0.93

 jail

0.91

 sleep

0.90

 hell

0.77

 grips

0.77

 warp

0.75

bat

0.75

Activations Density 0.119%

instances of the word "to" followed by a verb

instances of the word "to" indicating directional action or intentions

No Comments

No Known Activations