INDEX

Explanations

phrases related to action and doing something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "as"

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

Ire

-0.76

âĹ¼

-0.75

osc

-0.69

å½

-0.68

Redditor

-0.64

Secondly

-0.63

"></

-0.63

ãĥ¼ãĥĨ

-0.62

qv

-0.61

Category

-0.60

POSITIVE LOGITS

pires

0.98

pects

0.93

piring

0.93

pired

0.91

pire

0.89

phy

0.86

 soon

0.85

semb

0.85

semble

0.84

part

0.84

Activations Density 0.069%

phrases related to action and doing something

instances of the word "as"

No Comments

No Known Activations