INDEX

Explanations

phrases related to decision-making or action

oai_token-act-pair · gpt-3.5-turbo

modal verbs and expressions of intention or desire

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ãĤ¦ãĤ¹

-0.81

 fortun

-0.72

cedented

-0.72

ãĤ´ãĥ³

-0.69

built

-0.69

 Niet

-0.68

produced

-0.65

foreseen

-0.65

GoldMagikarp

-0.64

Introduced

-0.64

POSITIVE LOGITS

 give

1.13

 listen

1.11

 participate

1.10

 refuse

1.08

 make

1.07

 decide

1.04

 degrade

1.04

 send

1.04

 contribute

1.04

do

1.03

Activations Density 0.266%

phrases related to decision-making or action

modal verbs and expressions of intention or desire

No Comments

No Known Activations