INDEX

Explanations

verbs or phrases indicating a change or transition

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions, states, or conditions of entities

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

pite

-0.64

bury

-0.62

icol

-0.60

 Kills

-0.60

 Seah

-0.58

 Pearce

-0.58

 Smoking

-0.58

iful

-0.57

 afore

-0.56

 Thieves

-0.56

POSITIVE LOGITS

 nonetheless

0.72

ãĥĦ

0.70

ãĥĩãĤ£

0.70

 virtually

0.69

 nevertheless

0.69

 partName

0.66

 grav

0.65

 hardly

0.63

 compelled

0.63

ãĥķ

0.62

Activations Density 0.761%

verbs or phrases indicating a change or transition

phrases indicating actions, states, or conditions of entities

No Comments

No Known Activations