INDEX

Explanations

the word "probably" followed by a verb or adjective

oai_token-act-pair · gpt-3.5-turbo

the word "probably" and its associated concepts of uncertainty or conjecture

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

yip

-0.88

yss

-0.72

lete

-0.68

ieu

-0.67

ieve

-0.67

iya

-0.66

eli

-0.65

bender

-0.65

uctor

-0.65

letes

-0.64

POSITIVE LOGITS

 wouldn

1.09

 shouldn

1.08

won

1.02

 wont

0.97

 overest

0.94

 underestimate

0.88

 underest

0.88

won

0.86

 never

0.83

 weren

0.83

Activations Density 0.085%

the word "probably" followed by a verb or adjective

the word "probably" and its associated concepts of uncertainty or conjecture

No Comments

No Known Activations