INDEX

Explanations

references to actions taken against individuals or groups

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to actions taken against individuals, particularly in contexts of public service or legal matters

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ghost

-0.74

elman

-0.73

çĶ°

-0.68

aic

-0.67

maid

-0.67

wait

-0.67

ians

-0.66

bows

-0.63

batch

-0.63

Enlarge

-0.63

POSITIVE LOGITS

 consideration

1.16

 existence

1.00

 contention

0.97

 circulation

0.93

 sight

0.86

 participation

0.79

 relevance

0.79

 view

0.78

 participating

0.78

 duty

0.75

Activations Density 0.178%

references to actions taken against individuals or groups

references to actions taken against individuals, particularly in contexts of public service or legal matters

No Comments

No Known Activations