INDEX

Explanations

proper nouns related to historical figures, locations, and political entities

oai_token-act-pair · gpt-3.5-turbo

terms related to political and military entities or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

quickShipAvailable

-0.62

osponsors

-0.60

 Boko

-0.58

 bicycl

-0.56

hillary

-0.54

ospons

-0.54

axter

-0.54

çĶ°

-0.53

 discounts

-0.53

 Samar

-0.51

POSITIVE LOGITS

 universe

1.04

 ecosystem

0.97

 throne

0.96

iverse

0.94

 sphere

0.92

 cosmos

0.91

 realm

0.90

 system

0.89

 kingdom

0.88

 treasury

0.88

Activations Density 0.586%

proper nouns related to historical figures, locations, and political entities

terms related to political and military entities or situations

No Comments

No Known Activations