INDEX

Explanations

phrases related to positions or statuses in a specific context

oai_token-act-pair · gpt-3.5-turbo

phrases related to being in a specific state or condition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

aber

-0.66

 Adin

-0.65

anders

-0.65

arus

-0.63

 Ashton

-0.63

 Cutter

-0.61

 Ethernet

-0.59

ounds

-0.59

RX

-0.59

 hairst

-0.58

POSITIVE LOGITS

 financially

1.03

 situation

0.81

mma

0.81

ioned

0.79

pared

0.77

 circumstance

0.77

ï

0.77

ð

0.75

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

0.74

 psychologically

0.69

Activations Density 0.087%

phrases related to positions or statuses in a specific context

phrases related to being in a specific state or condition

No Comments

No Known Activations