INDEX

Explanations

personal names or titles that include "Di" followed by a number

oai_token-act-pair · gpt-3.5-turbo

references to the term "Di" associated with various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

manship

-0.92

sburgh

-0.85

éĹĺ

-0.78

nesday

-0.77

ãĥĦ

-0.75

Reviewer

-0.73

Thumbnail

-0.70

 firsthand

-0.69

erous

-0.67

*/(

-0.67

POSITIVE LOGITS

abol

1.30

abolic

1.22

Di

1.12

urnal

1.07

agram

1.04

ocese

1.01

agon

1.01

pora

0.99

agonal

0.95

ogo

0.93

Activations Density 0.004%

personal names or titles that include "Di" followed by a number

references to the term "Di" associated with various contexts

No Comments

No Known Activations