INDEX

Explanations

instances of the word "thinking" followed by a number

oai_token-act-pair · gpt-3.5-turbo

instances of the word "thinking."

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

çĦ

-0.78

feeding

-0.64

 Wrestling

-0.64

CBC

-0.63

Ann

-0.63

any

-0.61

videos

-0.60

clad

-0.60

 Videos

-0.59

owship

-0.59

POSITIVE LOGITS

 aloud

0.84

 provoking

0.81

sonian

0.80

cient

0.78

 about

0.74

eteen

0.72

ortment

0.70

lass

0.70

inery

0.69

 strategically

0.68

Activations Density 0.033%

instances of the word "thinking" followed by a number

instances of the word "thinking."

No Comments

No Known Activations