INDEX

Explanations

words containing the sub-string "onder"

oai_token-act-pair · gpt-3.5-turbo

words related to wonder or curiosity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 increment

-0.79

©¶æ

-0.75

fixes

-0.65

 crawl

-0.64

Ĺ

-0.64

 clamp

-0.63

Æ

-0.62

 laborers

-0.61

 projected

-0.59

 overrun

-0.59

POSITIVE LOGITS

lust

0.94

ful

0.94

mare

0.89

ance

0.85

oga

0.85

ãĤ¤ãĥĪ

0.83

FUL

0.83

lich

0.82

agher

0.82

hoe

0.82

Activations Density 0.019%

words containing the sub-string "onder"

words related to wonder or curiosity

No Comments

No Known Activations