INDEX

Explanations

statements indicating belief or disbelief in specific information

oai_token-act-pair · gpt-3.5-turbo

phrases expressing belief or skepticism

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

ific

-0.80

xes

-0.74

issance

-0.69

quit

-0.66

rection

-0.65

udo

-0.63

Thumbnail

-0.63

 wast

-0.62

backer

-0.62

nesses

-0.62

POSITIVE LOGITS

eele

0.71

 passionately

0.67

ASC

0.64

 Hier

0.63

REL

0.62

Orig

0.61

================================================================

0.61

 sincerity

0.61

 longevity

0.58

apesh

0.58

Activations Density 0.175%

statements indicating belief or disbelief in specific information

phrases expressing belief or skepticism

No Comments

No Known Activations