INDEX

Explanations

names of locations and titles

oai_token-act-pair · gpt-3.5-turbo

abbreviations or single-letter designations that represent names, terms, or entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

é¾įå¥ĳå£«

-0.68

staking

-0.68

DOE

-0.67

theless

-0.66

Pwr

-0.64

stellar

-0.63

shapeshifter

-0.63

 correspondence

-0.63

 trough

-0.61

 Sina

-0.59

POSITIVE LOGITS

ombat

0.91

isan

0.90

ooter

0.88

ox

0.84

ashes

0.84

itions

0.82

ancers

0.82

vironments

0.82

asions

0.79

ancer

0.78

Activations Density 0.326%

names of locations and titles

abbreviations or single-letter designations that represent names, terms, or entities

No Comments

No Known Activations