INDEX

Explanations

mentions of the word "for" occurring in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to the word "For."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

inese

-0.65

orically

-0.60

upon

-0.59

Jr

-0.59

mare

-0.58

 Burgess

-0.56

vom

-0.55

âĶ

-0.54

-0.53

icago

-0.53

POSITIVE LOGITS

bidden

1.50

gotten

1.48

geries

1.19

gery

1.19

ked

1.11

 example

1.08

gettable

1.06

 instance

1.02

cers

1.02

cing

1.01

Activations Density 0.136%

mentions of the word "for" occurring in various contexts

references to the word "For."

No Comments

No Known Activations