INDEX

Explanations

words related to notable figures (e.g., names of people)

oai_token-act-pair · gpt-3.5-turbo

Notable names or people mentioned in the text

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

iencies

-0.76

geries

-0.72

antine

-0.63

 overlap

-0.61

 locality

-0.60

orship

-0.59

isodes

-0.59

arial

-0.59

 Binding

-0.57

 parachute

-0.57

POSITIVE LOGITS

 replied

1.30

 exclaimed

1.24

 joked

1.20

 said

1.17

 tweeted

1.11

 explained

1.11

 remarked

1.11

âĢ

1.06

(@

1.06

 told

1.05

Activations Density 0.185%

words related to notable figures (e.g., names of people)

Notable names or people mentioned in the text

No Comments

No Known Activations