INDEX

Explanations

questions posed by 'Do we' or 'Do you'

oai_token-act-pair · gpt-3.5-turbo

questions about personal desires and collective actions

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

edIn

-0.69

 Flavoring

-0.68

itized

-0.67

 Vict

-0.65

ãĤ¦ãĤ¹

-0.65

 Conversation

-0.64

 srfAttach

-0.64

 Awakening

-0.64

uces

-0.64

 Integrity

-0.64

POSITIVE LOGITS

 intend

0.93

 think

0.92

 deserve

0.90

owe

0.88

 presume

0.87

 imply

0.86

 reckon

0.86

 look

0.85

 derive

0.85

 follow

0.85

Activations Density 0.085%

questions posed by 'Do we' or 'Do you'

questions about personal desires and collective actions

No Comments

No Known Activations