INDEX

Explanations

phrases related to educational institutions such as colleges and universities

oai_token-act-pair · gpt-3.5-turbo

mentions of educational institutions, specifically schools and universities

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

htaking

-0.73

Ô

-0.73

ravings

-0.72

hots

-0.71

chel

-0.70

cious

-0.68

NetMessage

-0.67

tml

-0.65

obyl

-0.65

aphael

-0.65

POSITIVE LOGITS

 itself

0.92

's

0.87

wide

0.83

 cafeteria

0.83

 logo

0.82

 superintendent

0.77

 hierarchy

0.75

 inspector

0.75

 headquarters

0.74

 councillor

0.74

Activations Density 0.270%

phrases related to educational institutions such as colleges and universities

mentions of educational institutions, specifically schools and universities

No Comments

No Known Activations