INDEX

Explanations

actions related to monitoring or inspecting something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "check" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Ã±a

-0.75

xual

-0.71

theless

-0.67

é¾įå

-0.66

 incent

-0.65

ufact

-0.65

usable

-0.65

SAY

-0.62

ãĥĢ

-0.62

åħī

-0.62

POSITIVE LOGITS

mate

1.05

 whether

1.01

boxes

0.92

lists

0.91

IDs

0.90

points

0.76

 boxes

0.75

out

0.74

 periodically

0.72

box

0.71

Activations Density 0.033%

actions related to monitoring or inspecting something

instances of the word "check" in various contexts

No Comments

No Known Activations