INDEX

Explanations

words that end with 'rem'

oai_token-act-pair · gpt-3.5-turbo

terms related to extremism and extreme behaviors

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Ao

-0.66

Ack

-0.64

 behold

-0.63

 sadd

-0.63

 steroids

-0.60

 Labrador

-0.59

SB

-0.58

 Typh

-0.57

 fetch

-0.57

Tec

-0.57

POSITIVE LOGITS

antle

1.30

phasis

1.10

lins

1.06

arkable

1.04

acists

1.02

acy

1.01

achine

1.01

obile

1.00

nant

1.00

ilitary

0.97

Activations Density 0.030%

words that end with 'rem'

terms related to extremism and extreme behaviors

No Comments

No Known Activations