INDEX

Explanations

phrases related to criticism and negative evaluations

oai_token-act-pair · gpt-3.5-turbo

negative assessments or critiques of performances and events

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sidx

-0.78

Allah

-0.72

yip

-0.71

CHAT

-0.70

 concess

-0.70

 maxim

-0.69

xus

-0.68

illance

-0.66

owers

-0.65

ocrin

-0.65

POSITIVE LOGITS

 lackluster

1.18

 sluggish

1.10

 disappointing

1.08

 mediocre

1.07

luster

1.05

 dismal

1.03

 disappoint

1.02

 plagued

1.02

 Worse

1.00

 disappointment

1.00

Activations Density 0.851%

phrases related to criticism and negative evaluations

negative assessments or critiques of performances and events

No Comments

No Known Activations