INDEX

Explanations

words related to policies and their impact

oai_token-act-pair · gpt-3.5-turbo

phrases detailing detrimental policies and their impacts

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

odiac

-0.75

rete

-0.71

raq

-0.70

Sync

-0.69

 Beaver

-0.69

reciation

-0.68

Pixel

-0.65

Registered

-0.62

atha

-0.62

 Feather

-0.61

POSITIVE LOGITS

 harms

1.53

 exacerbate

1.51

 undermine

1.49

 undermines

1.48

 exacerb

1.45

 impover

1.41

 devast

1.36

 undermined

1.35

 jeopard

1.35

 adversely

1.33

Activations Density 0.467%

words related to policies and their impact

phrases detailing detrimental policies and their impacts

No Comments

No Known Activations