INDEX

Explanations

years from 2011 to 2013

oai_token-act-pair · gpt-3.5-turbo

instances of specific years mentioned in the text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

heet

-0.78

izers

-0.77

itionally

-0.73

izer

-0.73

ierrez

-0.71

ishing

-0.71

abulary

-0.69

ishes

-0.67

urses

-0.67

itions

-0.66

POSITIVE LOGITS

å¹

0.89

-'

0.86

 onwards

0.81

election

0.74

bard

0.68

ãĤº

0.67

MY

0.64

0.62

 onward

0.61

å¤©

0.60

Activations Density 0.081%

years from 2011 to 2013

instances of specific years mentioned in the text

No Comments

No Known Activations