INDEX

Explanations

proper nouns related to people's names

oai_token-act-pair · gpt-3.5-turbo

occurrences of the substring "Re" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ãĥ¼ãĥĨãĤ£

-0.78

 Pebble

-0.75

åŃ

-0.69

DragonMagazine

-0.66

 à¨

-0.66

 Korra

-0.65

£ı

-0.65

 Northern

-0.64

 Nieto

-0.63

 Anchorage

-0.62

POSITIVE LOGITS

levant

1.24

becca

1.14

asonable

1.12

verend

1.12

usable

1.11

ceive

1.10

ceived

1.09

actor

1.08

versible

1.05

claim

1.04

Activations Density 0.026%

proper nouns related to people's names

occurrences of the substring "Re" in various contexts

No Comments

No Known Activations