INDEX

Explanations

expressions of uncertainty or concern

oai_token-act-pair · gpt-3.5-turbo

modal verbs indicating possibility or potential actions

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

roads

-0.68

maximum

-0.66

answered

-0.66

Dur

-0.66

fixes

-0.65

DAQ

-0.64

building

-0.63

Crunch

-0.63

 Bris

-0.61

 Pric

-0.60

POSITIVE LOGITS

 perceive

1.26

 violate

1.22

 inadvertently

1.21

be

1.21

 behave

1.20

 unwittingly

1.18

 suffer

1.17

 succumb

1.14

 misuse

1.14

 mistakenly

1.13

Activations Density 0.308%

expressions of uncertainty or concern

modal verbs indicating possibility or potential actions

No Comments

No Known Activations