INDEX

Explanations

phrases related to locations or events

oai_token-act-pair · gpt-3.5-turbo

numerical references or data related to rankings and statistics

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hement

-0.84

 eleph

-0.73

abouts

-0.67

aturdays

-0.64

 tremend

-0.62

ilipp

-0.61

 Learns

-0.59

 citiz

-0.59

 thieves

-0.57

 thous

-0.57

POSITIVE LOGITS

↵

0.85

Alright

0.73

Below

0.73

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.71

Spoiler

0.70

Course

0.65

Currently

0.64

-------

0.64

ÂłÂłÂłÂłÂłÂłÂłÂł

0.63

Similar

0.63

Activations Density 0.180%

phrases related to locations or events

numerical references or data related to rankings and statistics

No Comments

No Known Activations