INDEX

Explanations

the word "necessarily" occurring in sentences

oai_token-act-pair · gpt-3.5-turbo

expressions of uncertainty or qualifiers indicating that something is not always the case

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

banks

-0.95

iry

-0.81

ERAL

-0.74

RBI

-0.72

tein

-0.71

lace

-0.71

ults

-0.70

soon

-0.70

lain

-0.69

 Hunting

-0.68

POSITIVE LOGITS

 necessarily

1.02

 necess

0.99

 identifiable

0.87

 materially

0.85

necess

0.83

icably

0.80

 guiActiveUn

0.80

 conform

0.80

 detract

0.79

 obliged

0.78

Activations Density 0.020%

the word "necessarily" occurring in sentences

expressions of uncertainty or qualifiers indicating that something is not always the case

No Comments

No Known Activations