INDEX

Explanations

the word "only" followed by a noun or pronoun, suggesting a unique or exclusive situation

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rers

-0.70

ayne

-0.70

urchase

-0.67

 respectively

-0.67

ullah

-0.66

arbon

-0.65

ulk

-0.65

hops

-0.65

wash

-0.64

ensibly

-0.64

POSITIVE LOGITS

 thing

0.84

 culprit

0.83

 beneficiary

0.75

 troubled

0.74

 frontier

0.73

pmwiki

0.73

 contender

0.71

 phenomenon

0.70

 accol

0.69

 aspect

0.69

Activations Density 0.048%

the word "only" followed by a noun or pronoun, suggesting a unique or exclusive situation

No Comments

No Known Activations