INDEX

Explanations

significant years

oai_token-act-pair · gpt-3.5-turbo

references to specific years, particularly 1993 and 1994

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

Tu

-0.75

tro

-0.69

holder

-0.66

strings

-0.65

lying

-0.64

ledged

-0.64

edge

-0.64

hang

-0.63

sing

-0.62

hed

-0.62

POSITIVE LOGITS

ĸļ

0.82

uncture

0.79

-'

0.74

achine

0.74

ovies

0.71

å¹

0.70

kHz

0.68

theless

0.68

ixties

0.67

ohydrate

0.66

Activations Density 0.024%

significant years

references to specific years, particularly 1993 and 1994

No Comments

No Known Activations