INDEX

Explanations

phrases related to expressing uncertainty or making comparisons

oai_token-act-pair · gpt-3.5-turbo

contractions and repetitions of the verb "do."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ret

-0.70

Ñĥ

-0.68

ardless

-0.67

istration

-0.66

rolog

-0.65

araoh

-0.65

rum

-0.64

ctica

-0.64

andering

-0.63

Murray

-0.63

POSITIVE LOGITS

 Called

1.13

Ago

1.11

 Doesn

1.08

 Makes

1.08

 Them

1.07

 Cause

1.06

 Wrong

1.05

 Gets

1.04

 Definition

1.03

 Advice

1.02

Activations Density 0.082%

phrases related to expressing uncertainty or making comparisons

contractions and repetitions of the verb "do."

No Comments

No Known Activations