INDEX

Explanations

words related to evaluation, analysis, and usage of various things such as information, behaviors, and architecture

oai_token-act-pair · gpt-3.5-turbo

the word "by" appearing frequently in various 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

 resil

-0.70

LO

-0.67

ounter

-0.66

BSD

-0.64

wine

-0.64

bia

-0.63

çĶ°

-0.62

abi

-0.62

redits

-0.60

ensions

-0.60

POSITIVE LOGITS

products

1.03

 virtue

0.95

 passers

0.86

product

0.79

 anyone

0.79

 Europeans

0.79

 outsiders

0.76

 humankind

0.76

 everyone

0.76

 mankind

0.74

Activations Density 0.146%

words related to evaluation, analysis, and usage of various things such as information, behaviors, and architecture

the word "by" appearing frequently in various contexts

No Comments

No Known Activations