INDEX

Explanations

names and titles related to a variety of activities or roles

oai_token-act-pair · gpt-3.5-turbo

instances of the letter 'n' and variations of its representation in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ospons

-0.67

å°Ĩ

-0.60

Ń·

-0.60

 Ples

-0.60

 Krish

-0.59

cair

-0.59

-0.58

ayson

-0.58

 Phys

-0.58

POSITIVE LOGITS

-'

0.75

iverse

0.73

inion

0.71

ribly

0.67

FTWARE

0.65

FORMATION

0.65

llor

0.63

ACTED

0.63

raq

0.62

urai

0.61

Activations Density 0.108%

names and titles related to a variety of activities or roles

instances of the letter 'n' and variations of its representation in words

No Comments

No Known Activations