INDEX

Explanations

locations, particularly places with the prefix "Om"

oai_token-act-pair · gpt-3.5-turbo

geographical locations or references related to specific events or people

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

essors

-0.85

vous

-0.81

ãĥ¼ãĥĨãĤ£

-0.69

ãģ¦

-0.68

 Chapman

-0.67

Union

-0.64

 barriers

-0.63

=-=-=-=-

-0.62

Ð¾Ð

-0.62

å¥

-0.62

POSITIVE LOGITS

oland

1.10

yssey

1.03

essage

1.02

osc

0.98

olen

0.98

pora

0.96

oga

0.96

ortal

0.95

oms

0.94

omach

0.92

Activations Density 0.018%

locations, particularly places with the prefix "Om"

geographical locations or references related to specific events or people

No Comments

No Known Activations