INDEX

Explanations

phrases related to enforcement or intervention

oai_token-act-pair · gpt-3.5-turbo

phrases related to actions of diminishing or removing something

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

é¾įå

-0.63

 behalf

-0.61

åĪ

-0.55

å¹

-0.55

pron

-0.54

 bandwagon

-0.53

ãĥ¼ãĥĨãĤ£

-0.52

æ°

-0.51

adi

-0.50

gd

-0.50

POSITIVE LOGITS

 upon

1.62

on

1.45

upon

1.17

ON

1.17

On

1.12

On

1.09

onite

1.07

on

1.04

 Upon

1.01

 onto

0.96

Activations Density 0.448%

phrases related to enforcement or intervention

phrases related to actions of diminishing or removing something

No Comments

No Known Activations