INDEX

Explanations

phrases indicating capability or potential

oai_token-act-pair · gpt-3.5-turbo

terms related to the ability or capacity of objects or entities

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

cot

-0.70

cott

-0.69

meric

-0.69

chet

-0.65

cation

-0.64

mol

-0.63

Moz

-0.63

bard

-0.63

WAR

-0.63

ey

-0.62

POSITIVE LOGITS

bodied

0.86

nesses

0.82

¶ħ

0.78

 accompl

0.74

 impairment

0.73

 imag

0.72

 enough

0.70

icient

0.70

 feats

0.69

ioned

0.67

Activations Density 0.021%

phrases indicating capability or potential

terms related to the ability or capacity of objects or entities

No Comments

No Known Activations