INDEX

Explanations

terms and phrases related to determining or identifying something

oai_token-act-pair · gpt-3.5-turbo

phrases related to the concept of determining information or conditions

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

ãģĭ

-0.71

amon

-0.68

ãĤ´ãĥ³

-0.68

talking

-0.68

 Ibid

-0.68

Torrent

-0.66

ãģĨ

-0.64

raid

-0.63

ublic

-0.63

seys

-0.62

POSITIVE LOGITS

 whether

1.40

whether

1.04

if

0.97

 eligibility

0.96

how

0.96

 Whether

0.93

ially

0.93

 feasibility

0.93

 validity

0.92

 paternity

0.87

Activations Density 0.054%

terms and phrases related to determining or identifying something

phrases related to the concept of determining information or conditions

No Comments

No Known Activations