INDEX

Explanations

verbs related to asking or urging someone to take action

oai_token-act-pair · gpt-3.5-turbo

instances of the word "called" referring to requests or demands made by individuals or entities

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

bilt

-0.89

bourne

-0.72

 istg

-0.68

inth

-0.68

————

-0.68

olitics

-0.67

=-=-

-0.67

 ensu

-0.66

âķĲ

-0.65

Admin

-0.64

POSITIVE LOGITS

0.83

call

0.82

 bluff

0.81

 calling

0.81

igraph

0.80

calling

0.80

 Calling

0.80

phas

0.79

 Call

0.78

 upon

0.77

Activations Density 0.056%

verbs related to asking or urging someone to take action

instances of the word "called" referring to requests or demands made by individuals or entities

No Comments

No Known Activations