INDEX

Explanations

proper nouns, particularly related to individuals or places

oai_token-act-pair · gpt-3.5-turbo

terms related to spiritual leaders or gurus

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

iaries

-0.72

antine

-0.72

 skelet

-0.71

itent

-0.69

stros

-0.64

 selves

-0.63

ellen

-0.62

 histories

-0.61

 Winchester

-0.61

 corrid

-0.60

POSITIVE LOGITS

uru

1.10

hod

1.02

awei

0.93

isu

0.90

vana

0.87

umbai

0.86

ksh

0.83

geon

0.82

asu

0.80

ashtra

0.79

Activations Density 0.014%

proper nouns, particularly related to individuals or places

terms related to spiritual leaders or gurus

No Comments

No Known Activations