INDEX

Explanations

words related to quests or missions

oai_token-act-pair · gpt-3.5-turbo

the recurring theme of quests or searches for various goals or outcomes

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

quished

-0.66

 abusive

-0.64

avin

-0.61

PRES

-0.60

AGE

-0.60

 actu

-0.57

AGES

-0.57

sembly

-0.57

 Mock

-0.56

amera

-0.55

POSITIVE LOGITS

ioned

1.17

naire

1.03

naires

1.03

ogue

0.87

love

0.81

finder

0.78

uria

0.78

Â¯Â¯Â¯Â¯

0.74

lon

0.73

ues

0.73

Activations Density 0.018%

words related to quests or missions

the recurring theme of quests or searches for various goals or outcomes

No Comments

No Known Activations