INDEX

Explanations

references or mentions of specific entities or terms

oai_token-act-pair · gpt-3.5-turbo

frequent occurrences of the word "the" and related articles

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

iste

-0.85

pta

-0.78

paio

-0.74

awa

-0.70

 vulner

-0.68

âĢķ

-0.68

ãĤŃ

-0.66

bet

-0.65

ãĥ¼ãĥĨ

-0.64

 accordingly

-0.63

POSITIVE LOGITS

osures

0.73

 magnitude

0.70

 Americas

0.65

 Ages

0.59

ounding

0.59

 nutshell

0.57

ICH

0.57

 origin

0.55

 foregoing

0.55

 Devices

0.55

Activations Density 0.123%

references or mentions of specific entities or terms

frequent occurrences of the word "the" and related articles

No Comments

No Known Activations