INDEX

Explanations

the word "premise" in various contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "premise."

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

sung

-0.93

vor

-0.87

eday

-0.82

icer

-0.72

eder

-0.66

vals

-0.65

ese

-0.64

NetMessage

-0.64

pedia

-0.63

erers

-0.63

POSITIVE LOGITS

 premise

1.21

 premises

0.88

REC

0.70

IRE

0.69

 proposition

0.67

itial

0.66

 Budapest

0.66

OUND

0.65

ulhu

0.65

ALLY

0.64

Activations Density 0.019%

the word "premise" in various contexts

instances of the word "premise."

No Comments

No Known Activations