INDEX

Explanations

phrases indicating something or someone is a good choice or desirable

oai_token-act-pair · gpt-3.5-turbo

the word "as" used in various contexts and comparisons

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

kefeller

-0.72

eor

-0.61

eyes

-0.60

GOP

-0.60

çļĦ

-0.58

éĽ

-0.57

efe

-0.57

iery

-0.57

rir

-0.57

èĪ

-0.57

POSITIVE LOGITS

 evidenced

1.36

 well

1.28

well

1.23

 opposed

1.18

far

0.94

 regards

0.94

 they

0.94

 soon

0.92

 usual

0.91

ide

0.90

Activations Density 0.100%

phrases indicating something or someone is a good choice or desirable

the word "as" used in various contexts and comparisons

No Comments

No Known Activations