INDEX

Explanations

proper nouns related to locations and people

oai_token-act-pair · gpt-3.5-turbo

phrases that correspond to geographical locations and notable entities

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

•EggList

No Comments

Negative Logits

 Fairfax

-0.82

 âĹı

-0.82

ãĤ¼ãĤ¦ãĤ¹

-0.75

 Waste

-0.72

 Arlington

-0.71

-0.70

 Clemson

-0.69

 Wast

-0.69

 privat

-0.69

POSITIVE LOGITS

1.38

1.33

jer

1.17

Js

1.17

JD

1.16

jit

1.15

ij

1.12

jo

1.12

ji

1.10

je

1.10

Activations Density 0.153%

proper nouns related to locations and people

phrases that correspond to geographical locations and notable entities

No Comments

No Known Activations