INDEX

Explanations

phrases related to making things easier or harder

oai_token-act-pair · gpt-3.5-turbo

phrases related to the ease or difficulty of actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

abase

-0.67

 Words

-0.65

Stars

-0.65

utsche

-0.64

!/

-0.64

ODY

-0.63

 Variant

-0.61

chn

-0.61

 Bite

-0.61

 chang

-0.61

POSITIVE LOGITS

imaru

0.75

itary

0.73

Reilly

0.72

enged

0.67

 rout

0.67

aneously

0.65

safe

0.63

motion

0.62

cone

0.62

encia

0.62

Activations Density 0.059%

phrases related to making things easier or harder

phrases related to the ease or difficulty of actions or processes

No Comments

No Known Activations