INDEX

Explanations

instances where someone steps in to take action or make a statement

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "to."

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

interstitial

-0.93

TION

-0.70

icularly

-0.70

ivity

-0.67

iveness

-0.64

pointers

-0.64

htaking

-0.63

 Continued

-0.62

ient

-0.60

packing

-0.60

POSITIVE LOGITS

 defend

1.25

 compensate

1.22

 reinforce

1.22

 lend

1.18

 assist

1.16

 replace

1.14

 relieve

1.14

 participate

1.14

 fill

1.14

 contribute

1.13

Activations Density 0.141%

instances where someone steps in to take action or make a statement

occurrences of the word "to."

No Comments

No Known Activations