INDEX

Explanations

the word "series" followed by a noun

oai_token-act-pair · gpt-3.5-turbo

references to a sequence or collection of related items or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

lesi

-0.75

nai

-0.74

ossier

-0.73

lite

-0.73

illusion

-0.69

rella

-0.69

zan

-0.68

lean

-0.67

ryn

-0.67

arching

-0.65

POSITIVE LOGITS

manship

0.86

ãĥĺ

0.83

hews

0.77

ĸļ

0.77

 Tactics

0.73

series

0.71

pread

0.69

 finale

0.68

 è£ıè¦ļéĨĴ

0.67

Timeout

0.66

Activations Density 0.024%

the word "series" followed by a noun

references to a sequence or collection of related items or events

No Comments

No Known Activations