INDEX

Explanations

phrases related to critical commentary, reviews, or assessments

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

nces

-0.74

é¾į

-0.71

 Ancients

-0.70

doms

-0.67

bats

-0.66

ICES

-0.65

 Seasons

-0.64

iencies

-0.61

itudes

-0.60

ousands

-0.60

POSITIVE LOGITS

 breaker

0.84

 affair

0.81

 occurrence

0.79

 thing

0.78

 piece

0.78

brainer

0.77

 filler

0.76

 violation

0.76

 gimmick

0.76

 accomplishment

0.76

Activations Density 0.717%

phrases related to critical commentary, reviews, or assessments

No Comments

No Known Activations