INDEX

Explanations

phrases related to caution or warning against certain actions

oai_token-act-pair · gpt-3.5-turbo

references to quantity and recommendations regarding action

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

 exhibited

-0.49

 SetFontSize

-0.48

smanship

-0.48

 premiered

-0.47

encer

-0.47

encers

-0.47

 Photograph

-0.46

çīĪ

-0.45

 Bene

-0.45

 Symphony

-0.45

POSITIVE LOGITS

livion

0.60

 wilderness

0.51

OSP

0.48

 vulner

0.48

 recess

0.47

 reckoning

0.47

 defenses

0.47

 ledge

0.45

 difficulties

0.45

tackle

0.45

Activations Density 1.699%

phrases related to caution or warning against certain actions

references to quantity and recommendations regarding action

No Comments

No Known Activations