INDEX

Explanations

language related to implication or suggestion, especially with negative connotations

oai_token-act-pair · gpt-3.5-turbo

instances of the term "impl" or related phrases indicating implementation or implication

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

Ĥ¬

-0.81

SEA

-0.77

grass

-0.76

chal

-0.73

flix

-0.72

¥µ

-0.68

PDATE

-0.67

hyde

-0.67

 Warwick

-0.65

boarding

-0.65

POSITIVE LOGITS

osion

1.41

oded

1.34

ausible

1.27

icating

1.21

icates

1.20

icate

1.18

anting

1.17

icit

1.16

oding

1.15

ications

1.08

Activations Density 0.011%

language related to implication or suggestion, especially with negative connotations

instances of the term "impl" or related phrases indicating implementation or implication

No Comments

No Known Activations