INDEX

Explanations

phrases related to valuable items or possessions

oai_token-act-pair · gpt-3.5-turbo

references to valuable tangible items and resources

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

erate

-0.86

 interven

-0.69

 theoret

-0.66

atalie

-0.65

Reply

-0.65

Reviewer

-0.64

âĸ¬

-0.64

Relations

-0.64

ALLY

-0.63

 advers

-0.62

POSITIVE LOGITS

mith

1.38

leeve

1.14

creen

1.09

ilver

1.06

poons

1.06

hift

1.04

pring

0.99

cape

0.98

hare

0.96

hops

0.95

Activations Density 0.427%

phrases related to valuable items or possessions

references to valuable tangible items and resources

No Comments

No Known Activations