INDEX

Explanations

phrases related to growth or increase

oai_token-act-pair · gpt-3.5-turbo

the word "in" occurring frequently within various contexts

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

killed

-0.73

awan

-0.66

HAEL

-0.65

PIN

-0.63

awei

-0.62

ens

-0.62

PASS

-0.61

%%

-0.61

wine

-0.60

=""

-0.57

POSITIVE LOGITS

roads

1.14

 accordance

1.12

 unison

1.11

clusions

1.09

 spite

1.08

 regards

1.07

 favor

1.07

 tandem

1.06

ordinate

1.06

 vitro

1.06

Activations Density 0.305%

phrases related to growth or increase

the word "in" occurring frequently within various contexts

No Comments

No Known Activations