INDEX

Explanations

organizations or departments and their actions or policies

oai_token-act-pair · gpt-3.5-turbo

repeated references to governmental or legal entities

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

 disadvant

-0.75

 smugglers

-0.71

 camouflage

-0.69

 readings

-0.68

jog

-0.67

 puff

-0.67

 reproduction

-0.67

 whiff

-0.67

 litter

-0.66

geries

-0.65

POSITIVE LOGITS

ï¸ı

1.20

ï¸

1.02

SHIP

0.91

taboola

0.84

iversary

0.83

WARE

0.81

tre

0.81

owners

0.81

worthiness

0.80

mand

0.80

Activations Density 0.398%

organizations or departments and their actions or policies

repeated references to governmental or legal entities

No Comments

No Known Activations