INDEX

Explanations

specific terms or phrases known by a particular name or title

oai_token-act-pair · gpt-3.5-turbo

references to specific entities or titles being introduced or defined

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tempted

-0.72

 convinced

-0.72

SPONSORED

-0.71

 furthermore

-0.70

laim

-0.68

 headquartered

-0.68

Joined

-0.68

olicy

-0.67

 inclined

-0.66

 horr

-0.64

POSITIVE LOGITS

phe

0.75

"#

0.73

sbm

0.73

0.72

 Handbook

0.71

 Excellence

0.71

ãĢİ

0.70

mother

0.69

 Gleaming

0.69

0.68

Activations Density 0.141%

specific terms or phrases known by a particular name or title

references to specific entities or titles being introduced or defined

No Comments

No Known Activations