INDEX

Explanations

structured and numbered text data, possibly related to member activities or profiles

oai_token-act-pair · gpt-3.5-turbo

structured data or metrics related to performance and membership in a system

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

eus

-0.90

£ı

-0.77

nep

-0.76

oÄŁ

-0.69

 whiff

-0.68

è£

-0.67

dfx

-0.64

Bey

-0.64

pse

-0.63

FOX

-0.63

POSITIVE LOGITS

Merit

0.94

 Posts

0.80

Thread

0.77

 Siege

0.73

irc

0.73

 Armory

0.73

swick

0.72

 Likes

0.72

 Login

0.72

essions

0.71

Activations Density 0.066%

structured and numbered text data, possibly related to member activities or profiles

structured data or metrics related to performance and membership in a system

No Comments

No Known Activations