INDEX

Explanations

proper nouns or names, specifically focusing on the name "Stra" in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to a specific name or entity, likely a journalist or publication

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

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.78

gdala

-0.77

flies

-0.76

 Ashton

-0.75

è¦ļéĨĴ

-0.74

eor

-0.68

 GOODMAN

-0.64

 Webster

-0.63

NCT

-0.62

orative

-0.61

POSITIVE LOGITS

pping

0.97

fing

0.94

ights

0.93

eties

0.91

pped

0.91

pper

0.89

 Stra

0.83

keye

0.83

qi

0.83

Stra

0.81

Activations Density 0.017%

proper nouns or names, specifically focusing on the name "Stra" in various contexts

references to a specific name or entity, likely a journalist or publication

No Comments

No Known Activations