INDEX

Explanations

words related to the author "Tol."

oai_token-act-pair · gpt-3.5-turbo

references to specific names or locations associated with a particular context

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

 Sandy

-0.80

çīĪ

-0.79

æŃ¦

-0.74

 Murdoch

-0.71

--------------------------------------------------------

-0.70

Ear

-0.69

 Hanson

-0.68

å£«

-0.67

 Brand

-0.67

lihood

-0.67

POSITIVE LOGITS

ological

1.10

ologically

1.02

ologic

1.00

ogen

0.97

uno

0.95

ologies

0.94

nil

0.88

ology

0.88

opl

0.87

okin

0.87

Activations Density 0.042%

words related to the author "Tol."

references to specific names or locations associated with a particular context

No Comments

No Known Activations