INDEX

Explanations

phrases prompting reader interaction or engagement

oai_token-act-pair · gpt-3.5-turbo

frequent occurrences of the word "you."

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

ruciating

-0.79

ospital

-0.71

apo

-0.70

uces

-0.69

images

-0.68

icent

-0.68

uction

-0.68

ape

-0.66

Pg

-0.66

Lago

-0.65

POSITIVE LOGITS

 guys

1.11

're

0.99

tub

0.96

 know

0.86

 prefer

0.86

 subscribed

0.85

 wanna

0.85

 want

0.84

 intend

0.84

've

0.84

Activations Density 0.054%

phrases prompting reader interaction or engagement

frequent occurrences of the word "you."

No Comments

No Known Activations