INDEX

Explanations

phrases related to feedback and opinions

oai_token-act-pair · gpt-3.5-turbo

phrases and words connected to approval and criticism

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

agate

-0.75

 planner

-0.71

 exercises

-0.68

 Parallel

-0.67

wine

-0.63

 supplements

-0.62

 specialization

-0.62

walking

-0.60

empl

-0.60

artifacts

-0.60

POSITIVE LOGITS

 audiences

1.25

 detractors

1.17

 onlook

1.15

 critics

1.14

 applause

1.06

 skeptics

1.04

 listeners

1.03

 viewers

1.01

 supporters

0.98

 disapproval

0.98

Activations Density 1.374%

phrases related to feedback and opinions

phrases and words connected to approval and criticism

No Comments

No Known Activations