INDEX

Explanations

phrases indicating availability or readiness for action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating something is available or being offered

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

cles

-0.70

ellen

-0.68

ACTED

-0.66

aan

-0.66

aghan

-0.66

sequently

-0.65

een

-0.64

leigh

-0.64

MJ

-0.63

ensen

-0.62

POSITIVE LOGITS

 instance

0.83

 example

0.81

gery

0.79

going

0.78

geries

0.78

 grabs

0.78

gotten

0.71

 emergencies

0.68

Ĥª

0.68

 starters

0.67

Activations Density 0.116%

phrases indicating availability or readiness for action

phrases indicating something is available or being offered

No Comments

No Known Activations