INDEX

Explanations

proper nouns related to locations, especially Washington, D.C

oai_token-act-pair · gpt-3.5-turbo

punctuations and formatting related to listings

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

 sentimental

-0.71

Reviewer

-0.71

 goodwill

-0.65

ãģĦ

-0.64

izations

-0.64

 purse

-0.63

 fins

-0.63

 gears

-0.62

 incomes

-0.62

ãĤ¿

-0.61

POSITIVE LOGITS

alus

0.93

rex

0.86

Gray

0.85

Va

0.83

 Eisenhower

0.80

mort

0.79

 Rockefeller

0.77

mun

0.74

 Dull

0.72

ember

0.71

Activations Density 0.050%

proper nouns related to locations, especially Washington, D.C

punctuations and formatting related to listings

No Comments

No Known Activations