INDEX

Explanations

verbs in the infinitive form

oai_token-act-pair · gpt-3.5-turbo

modal verbs indicating necessity or obligation

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Globe

-0.75

 Hamm

-0.67

Gw

-0.63

achu

-0.63

 Printing

-0.61

Zip

-0.61

 miscarriage

-0.60

lash

-0.60

Alm

-0.59

POSITIVE LOGITS

 careful

1.02

 reckoned

0.98

 accounted

0.96

 sacrificed

0.95

 rescued

0.93

 avoided

0.92

 dealt

0.91

 kept

0.91

 rebuilt

0.90

 understood

0.88

Activations Density 0.070%

verbs in the infinitive form

modal verbs indicating necessity or obligation

No Comments

No Known Activations