INDEX

Explanations

names of people or entities associated with a topic or event

oai_token-act-pair · gpt-3.5-turbo

names of authors and their publication years

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 comr

-0.71

?).

-0.67

 evidenced

-0.66

 haun

-0.65

?'"

-0.63

 marginally

-0.62

 compassion

-0.62

 steroids

-0.61

 persists

-0.61

 comeback

-0.61

POSITIVE LOGITS

 Published

1.21

 Posted

1.16

 Reporter

1.10

 Contribut

1.07

 Random

1.04

 SECTION

1.04

1.03

 âĢ¢

1.02

 Staff

1.00

 Updated

0.99

Activations Density 0.115%

names of people or entities associated with a topic or event

names of authors and their publication years

No Comments

No Known Activations