INDEX

Explanations

phrases related to specific names, specifically "Ved" and "Rig"

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

è¦ļéĨĴ

-0.84

isters

-0.74

hips

-0.74

icago

-0.73

éĹĺ

-0.67

iculty

-0.66

channelAvailability

-0.65

è£

-0.64

\/\/

-0.63

 VICE

-0.63

POSITIVE LOGITS

Rig

1.02

glers

0.94

orno

0.86

rupted

0.85

abase

0.83

Ved

0.78

ging

0.78

ged

0.73

rehend

0.70

gil

0.70

Activations Density 0.012%

phrases related to specific names, specifically "Ved" and "Rig"

proper nouns, particularly names and titles

No Comments

No Known Activations