INDEX

Explanations

the word 'Del' with varying degrees of activation

oai_token-act-pair · gpt-3.5-turbo

mentions of "Del" likely referring to a specific entity or brand

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

SHIP

-0.68

 altogether

-0.64

AFL

-0.62

 Miko

-0.61

 stand

-0.58

...]

-0.58

PsyNetMessage

-0.58

 fortune

-0.57

 fellowship

-0.57

 Authorization

-0.56

POSITIVE LOGITS

icate

1.40

ayed

1.40

usional

1.36

phi

1.28

ivered

1.25

iber

1.22

ivery

1.20

gado

1.20

ivering

1.19

aware

1.15

Activations Density 0.016%

the word 'Del' with varying degrees of activation

mentions of "Del" likely referring to a specific entity or brand

No Comments

No Known Activations