INDEX

Explanations

phrases related to physical conflict or altercations

oai_token-act-pair · gpt-3.5-turbo

actions and events related to confrontation or violence

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

 Inher

-0.71

etheless

-0.65

efeated

-0.65

 inher

-0.63

 bestselling

-0.63

 quant

-0.61

 dearly

-0.61

 prag

-0.61

 archives

-0.61

obyl

-0.60

POSITIVE LOGITS

 altercation

0.67

 startled

0.67

 disturbance

0.66

 gunfire

0.66

 grabbed

0.65

 agitated

0.63

Stop

0.62

 takedown

0.62

 overpowered

0.61

 yelling

0.61

Activations Density 0.872%

phrases related to physical conflict or altercations

actions and events related to confrontation or violence

No Comments

No Known Activations