INDEX

Explanations

political figures and events

oai_token-act-pair · gpt-3.5-turbo

references to significant incidents or investigations involving individuals in power

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

ãĤ¦ãĤ¹

-0.66

ãĥ´

-0.62

Brit

-0.61

 proverb

-0.61

ngth

-0.60

illions

-0.60

ictionary

-0.60

å¹

-0.60

Ĭ±

-0.59

anooga

-0.59

POSITIVE LOGITS

 embattled

0.59

 Titan

0.59

 cryptocurrency

0.58

UCS

0.57

 Fresno

0.57

 NYPD

0.56

 upcoming

0.56

 Albuquerque

0.56

PSU

0.55

MTA

0.55

Activations Density 1.269%

political figures and events

references to significant incidents or investigations involving individuals in power

No Comments

No Known Activations