INDEX

Explanations

the word "latter" in comparisons

oai_token-act-pair · gpt-3.5-turbo

the term "latter" and its contextual usage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

bons

-0.65

 Applications

-0.59

 every

-0.56

 anywhere

-0.54

 whats

-0.54

ards

-0.54

ored

-0.54

 anytime

-0.52

 where

-0.52

isd

-0.51

POSITIVE LOGITS

 latter

4.07

 aforementioned

1.33

theless

1.12

 former

1.11

 foregoing

1.08

 respectively

1.01

 remainder

1.00

 afore

0.97

 ensuing

0.94

 resultant

0.93

Activations Density 0.011%

the word "latter" in comparisons

the term "latter" and its contextual usage

No Comments

No Known Activations