INDEX

Explanations

phrases related to luring or baiting

oai_token-act-pair · gpt-3.5-turbo

language related to deception or entrapment

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

urity

-0.82

blance

-0.73

oret

-0.69

yrus

-0.69

eas

-0.69

olitan

-0.68

ppard

-0.67

eely

-0.66

ias

-0.65

iator

-0.65

POSITIVE LOGITS

 lure

1.08

 bait

1.03

ument

0.96

mong

0.90

glers

0.84

EStream

0.82

Wag

0.78

GGGGGGGG

0.76

crow

0.71

ãĥĦ

0.70

Activations Density 0.036%

phrases related to luring or baiting

language related to deception or entrapment

No Comments

No Known Activations