INDEX

Explanations

words related to news articles about various events and topics

oai_token-act-pair · gpt-3.5-turbo

phrases related to military and conflict situations

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etsk

-0.78

dyl

-0.74

auth

-0.70

ipl

-0.69

nob

-0.68

fal

-0.67

akings

-0.65

 scout

-0.65

 disarm

-0.63

naissance

-0.62

POSITIVE LOGITS

 Relationship

0.87

 Them

0.87

 Dating

0.85

 Himself

0.82

 Advantage

0.82

 Donation

0.80

 Tone

0.80

 Purpose

0.79

 Favor

0.79

 Guys

0.79

Activations Density 0.707%

words related to news articles about various events and topics

phrases related to military and conflict situations

No Comments

No Known Activations