INDEX

Explanations

references to physical or digital cards being used symbolically or literally in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to "card" in various contexts

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

¿½

-0.73

ItemTracker

-0.73

Kut

-0.73

 Fiction

-0.71

 Progress

-0.70

 Skydragon

-0.69

ithing

-0.68

%]

-0.64

 Siem

-0.64

é¾

-0.62

POSITIVE LOGITS

iologist

1.29

inals

1.27

igan

1.24

igans

1.16

card

1.01

iology

1.01

iovascular

0.99

holders

0.99

iop

0.97

Card

0.96

Activations Density 0.022%

references to physical or digital cards being used symbolically or literally in various contexts

references to "card" in various contexts

No Comments

No Known Activations