INDEX

Explanations

phrases related to general discussions or transitions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating conditional scenarios or alternatives

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

intent

-0.72

 necks

-0.71

usage

-0.70

mens

-0.68

 repositories

-0.67

bent

-0.66

mers

-0.66

coded

-0.66

bots

-0.65

bian

-0.62

POSITIVE LOGITS

 thereafter

0.81

 concede

0.79

................................................................

0.76

 except

0.72

 suffice

0.72

hew

0.72

[/

0.71

 else

0.71

 concludes

0.71

|--

0.70

Activations Density 0.046%

phrases related to general discussions or transitions

phrases indicating conditional scenarios or alternatives

No Comments

No Known Activations