INDEX

Explanations

names and titles of individuals or entities

oai_token-act-pair · gpt-3.5-turbo

proper nouns or names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 eleph

-0.76

 BALL

-0.72

 å¤

-0.67

VEN

-0.65

ĵĺ

-0.65

RED

-0.65

 Vert

-0.65

 Percy

-0.64

-0.63

©¶æ

-0.63

POSITIVE LOGITS

1.94

iologist

1.23

ei

1.18

iak

1.16

ih

1.10

iology

1.09

ii

1.08

iator

1.05

ie

0.99

Activations Density 0.100%

names and titles of individuals or entities

proper nouns or names

No Comments

No Known Activations