INDEX

Explanations

phrases related to tasks or activities that need to be accomplished

oai_token-act-pair · gpt-3.5-turbo

phrases related to past actions and their implications

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

actionDate

-0.74

culosis

-0.73

erva

-0.69

leading

-0.63

Camp

-0.63

 Hast

-0.61

erity

-0.60

Ring

-0.59

Shape

-0.59

rower

-0.59

POSITIVE LOGITS

 happening

1.18

 discussed

1.17

 implemented

1.14

 done

1.11

 included

1.10

 explored

1.10

 achievable

1.08

 mentioned

1.08

 incorporated

1.06

 communicated

1.06

Activations Density 0.305%

phrases related to tasks or activities that need to be accomplished

phrases related to past actions and their implications

No Comments

No Known Activations