INDEX

Explanations

breaking news

oai_token-act-pair · gpt-3.5-turbo

terms related to breaking news and alerts

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

utical

-0.72

HAHA

-0.71

sbm

-0.70

lac

-0.65

 phys

-0.64

oka

-0.63

gul

-0.63

wom

-0.62

 bargaining

-0.62

 builders

-0.62

POSITIVE LOGITS

Desk

0.77

 Courier

0.76

 Emails

0.75

ource

0.74

CNN

0.71

weekly

0.70

Brow

0.69

letter

0.69

ylon

0.68

alion

0.67

Activations Density 0.017%

breaking news

terms related to breaking news and alerts

No Comments

No Known Activations