INDEX

Explanations

instances where people encounter or bump into each other

oai_token-act-pair · gpt-3.5-turbo

the phrase "run into."

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

atts

-0.74

arming

-0.71

iev

-0.70

end

-0.69

lege

-0.67

caster

-0.67

hop

-0.66

cham

-0.66

lime

-0.66

edition

-0.65

POSITIVE LOGITS

ãĥķ

0.77

âĸĳâĸĳ

0.71

-+-+

0.69

ODE

0.66

ãĤ´ãĥ³

0.65

ãĤ¨ãĥ«

0.65

 attractions

0.62

ãĥīãĥ©ãĤ´ãĥ³

0.62

âĸĴ

0.62

OVA

0.61

Activations Density 0.027%

instances where people encounter or bump into each other

the phrase "run into."

No Comments

No Known Activations