INDEX

Explanations

words related to interrogation, investigation, and scandals, particularly those related to the term "interrogate."

oai_token-act-pair · gpt-3.5-turbo

words related to investigations and interferences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ongyang

-0.84

erest

-0.76

uden

-0.70

gran

-0.68

apest

-0.67

bery

-0.64

corn

-0.64

jri

-0.64

haps

-0.62

 crop

-0.62

POSITIVE LOGITS

enment

0.91

EMENT

0.86

 Lauder

0.84

ATIVE

0.82

ciating

0.81

ATION

0.80

achment

0.79

ATOR

0.77

ATIONS

0.76

MENT

0.76

Activations Density 0.182%

words related to interrogation, investigation, and scandals, particularly those related to the term "interrogate."

words related to investigations and interferences

No Comments

No Known Activations