INDEX

Explanations

points or arguments being made in a text

oai_token-act-pair · gpt-3.5-turbo

references to key points being made in discussions or arguments

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

DAQ

-0.86

reditary

-0.84

uthor

-0.84

 destro

-0.82

 notor

-0.80

eatures

-0.78

apons

-0.77

eco

-0.75

undai

-0.75

emale

-0.74

POSITIVE LOGITS

lessly

0.93

point

0.89

points

0.88

blank

0.82

posted

0.82

forward

0.82

lessness

0.77

 deduction

0.75

posts

0.75

iasis

0.74

Activations Density 0.033%

points or arguments being made in a text

references to key points being made in discussions or arguments

No Comments

No Known Activations