INDEX

Explanations

words related to actions or events that are obvious, clearly shown, or openly done

oai_token-act-pair · gpt-3.5-turbo

references to overt actions or visibility in contexts

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

nesota

-0.82

STON

-0.70

«ĺ

-0.70

llah

-0.67

haus

-0.66

å¥³

-0.57

osuke

-0.56

 respectfully

-0.56

SAR

-0.55

ISO

-0.55

POSITIVE LOGITS

orial

1.42

aken

1.26

ures

1.24

reatment

1.24

ure

1.19

uned

1.18

orical

1.15

aking

1.12

urned

1.05

uning

1.04

Activations Density 0.064%

words related to actions or events that are obvious, clearly shown, or openly done

references to overt actions or visibility in contexts

No Comments

No Known Activations