INDEX

Explanations

proper names, particularly the name "Lyell"

oai_token-act-pair · gpt-3.5-turbo

the repetition and mention of the name "Ly" or variations of it in different contexts

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

perture

-0.93

raints

-0.78

atchewan

-0.77

sburgh

-0.73

boards

-0.69

ORTS

-0.68

ardless

-0.67

arthy

-0.65

ajor

-0.64

insula

-0.64

POSITIVE LOGITS

rics

0.93

onel

0.85

Ly

0.84

nda

0.84

comed

0.80

mb

0.80

umen

0.77

hurst

0.76

rell

0.74

low

0.74

Activations Density 0.015%

proper names, particularly the name "Lyell"

the repetition and mention of the name "Ly" or variations of it in different contexts

No Comments

No Known Activations