INDEX

Explanations

personal pronouns and short verb phrases, potentially indicating dialogue

oai_token-act-pair · gpt-3.5-turbo

expressions of uncertainty and indecision in dialogue

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

uably

-0.55

ufact

-0.54

ordes

-0.52

 agric

-0.52

issance

-0.51

AMD

-0.50

 municip

-0.50

orsi

-0.50

tnc

-0.50

respective

-0.49

POSITIVE LOGITS

 fuckin

0.86

 fucking

0.81

uh

0.71

 fucked

0.70

 fuck

0.68

eeee

0.67

 gonna

0.67

 funny

0.66

 goddamn

0.65

 kinda

0.64

Activations Density 1.658%

personal pronouns and short verb phrases, potentially indicating dialogue

expressions of uncertainty and indecision in dialogue

No Comments

No Known Activations