INDEX

Explanations

phrases related to singular, specific instances or events

oai_token-act-pair · gpt-3.5-turbo

words related to exceptions, discontinuities, or unique instances

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

emetery

-0.65

usable

-0.63

anooga

-0.62

ï¿½ï¿½

-0.61

 Instr

-0.60

¹

-0.60

grave

-0.59

 tradem

-0.59

ª

-0.59

srf

-0.58

POSITIVE LOGITS

 underdog

0.67

 darling

0.64

 quiz

0.63

 charm

0.60

manship

0.58

coy

0.58

lihood

0.58

 pessim

0.58

ishly

0.57

 goodbye

0.57

Activations Density 0.447%

phrases related to singular, specific instances or events

words related to exceptions, discontinuities, or unique instances

No Comments

No Known Activations