INDEX

Explanations

instances of list items or options identified by commas followed by a numerical value indicating significance or priority

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly commas and parentheses

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

cellence

-0.67

rero

-0.66

è¦ļéĨĴ

-0.65

»Ĵ

-0.64

ãĤ¨ãĥ«

-0.64

formance

-0.61

rio

-0.59

ãĤº

-0.58

ensable

-0.57

ãĤ¦ãĤ¹

-0.55

POSITIVE LOGITS

 maybe

1.19

 depending

1.16

but

1.13

 perhaps

1.12

maybe

1.11

perhaps

1.09

or

1.07

 possibly

1.02

possibly

1.01

 Alternatively

1.01

Activations Density 0.447%

instances of list items or options identified by commas followed by a numerical value indicating significance or priority

punctuation marks, particularly commas and parentheses

No Comments

No Known Activations