INDEX

Explanations

specific years, likely related to events or documents published in those years

oai_token-act-pair · gpt-3.5-turbo

references to the year 2014

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

ishing

-0.97

isher

-0.82

ishes

-0.81

ype

-0.81

abilia

-0.80

itionally

-0.76

ishment

-0.75

ishers

-0.75

wo

-0.73

izing

-0.68

POSITIVE LOGITS

å¹

1.01

 onwards

0.79

assetsadobe

0.76

âĹ¼

0.76

etheus

0.71

 Sochi

0.69

âĢ³

0.68

episode

0.64

âĢ²

0.63

Activations Density 0.031%

specific years, likely related to events or documents published in those years

references to the year 2014

No Comments

No Known Activations