INDEX

Explanations

specific times and locations mentioned in the text

oai_token-act-pair · gpt-3.5-turbo

details related to specific incidents or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 haunted

-0.68

chwitz

-0.64

 traitor

-0.62

 exile

-0.62

 haunt

-0.61

 cures

-0.61

obyl

-0.60

ucket

-0.60

})

-0.59

letters

-0.59

POSITIVE LOGITS

 shortly

0.75

 according

0.68

 approximately

0.67

 southwest

0.67

 northwest

0.63

 located

0.62

 near

0.60

 Moderate

0.60

 northeast

0.58

endiary

0.57

Activations Density 0.418%

specific times and locations mentioned in the text

details related to specific incidents or events

No Comments

No Known Activations