INDEX

Explanations

ways or methods of doing something

oai_token-act-pair · gpt-3.5-turbo

phrases that describe actions or qualities in relation to "way" or "that."

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

RN

-0.69

PW

-0.68

NB

-0.67

 ..............

-0.65

FS

-0.64

 Sheet

-0.64

Ov

-0.64

Pog

-0.63

Ruk

-0.62

 Horizons

-0.62

POSITIVE LOGITS

 permitting

0.68

avior

0.66

 derog

0.66

forward

0.65

versible

0.65

apist

0.65

reflect

0.64

emis

0.63

 conduit

0.63

appropriate

0.63

Activations Density 0.204%

ways or methods of doing something

phrases that describe actions or qualities in relation to "way" or "that."

No Comments

No Known Activations