INDEX

Explanations

locations or specific positions within a setting

oai_token-act-pair · gpt-3.5-turbo

word forms that indicate existence or states of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

Enlarge

-0.70

buster

-0.69

chairs

-0.66

 didnt

-0.65

wired

-0.64

ascript

-0.62

 submar

-0.61

ples

-0.60

worthiness

-0.60

factor

-0.60

POSITIVE LOGITS

Sharp

0.66

 Glory

0.65

 Sharp

0.65

 Greenberg

0.64

 Zamb

0.64

 Gord

0.64

ports

0.63

gression

0.62

 myriad

0.62

USB

0.61

Activations Density 0.107%

locations or specific positions within a setting

word forms that indicate existence or states of being

No Comments

No Known Activations